- Email: [email protected]
Prevention of Illicit Drug Use Through a School-Based Program: Results of a Longitudinal, Cluster-Randomized Controlled Trial
Journal of Adolescent Health 56 (2015) 314e322
www.jahonline.org Original article
Prevention of Illicit Drug Use Through a School-Based Program: Results of a Longitudinal, Cluster-Randomized Controlled Trial Jong-Long Guo, Ph.D. a, Tzu-Chi Lee, Ph.D. b, Jung-Yu Liao, M.Ed. a, and Chiu-Mieh Huang, Ph.D. c, * a b c
Department of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan Department of Public Health, Kaohsiung Medical University, Kaohsiung, Taiwan School of Nursing, National Yang-Ming University, Taipei, Taiwan
Article history: Received October 16, 2013; Accepted December 1, 2014 Keywords: Illicit drug use prevention program; Booster session; Theory of planned behavior; Life skills; Perceived behavioral control; Behavioral intention
A B S T R A C T
Purpose: To evaluate the long-term effects of an illicit drug use prevention program for adolescents that integrates life skills into the theory of planned behavior. Methods: We conducted a cluster-randomized trial in which 24 participating schools were randomized to either an intervention group (12 schools, n ¼ 1,176 students) or a control group (12 schools, n ¼ 915 students). Participants were grade 7 students. The intervention comprised a main intervention of 10 sessions and two booster interventions. Booster 1 (four sessions) and booster 2 (two sessions) were performed at 6 months and 12 months, respectively, after completion of the main intervention. Assessments were made at baseline, after the main intervention, and after each booster session using specific questionnaires for measuring participants’ attitudes, subjective norms, perceived behavioral control, and life skills. Results: Retention rates were 71.9% (845/1,176) in the intervention group and 90.7% (830/915) in the control group after the 12-month follow-up. A significantly lower proportion of intervention group participants reported illicit drug use after the first and second booster sessions compared with control group participants (.1% vs. 1.7% and .2% vs. 1.7%, respectively; both p < .05). Attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention scores of the intervention group were significantly higher than those of control group after the first and second booster sessions (all p < .001), suggesting that intervention group students tended to avoid drug use. Conclusions: A drug use prevention program integrating life skills into the theory of planned behavior may be effective for reducing illicit drug use and improving planned behavior-related constructs in adolescents. Ó 2015 Society for Adolescent Health and Medicine. All rights reserved.
Illicit drug use can have a number of negative consequences, including physical and mental disorders, unemployment, accidents, suicide, violence, and a significant financial burden [1]. Adolescence Conflicts of Interest: The authors have no conflicts of interest to declare. * Address correspondence to: Chiu-Mieh Huang, Ph.D., School of Nursing, National Yang-Ming University, 155, Linong Street, Section 2, Taipei, Taiwan 11221. E-mail address: [email protected] (C.-M. Huang). 1054-139X/Ó 2015 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/j.jadohealth.2014.12.003
IMPLICATIONS AND CONTRIBUTION
An intervention that incorporated the theory of planned behavior and life skills helped to promote participants’ attitudes, subjective norms, perceived behavioral control, and behavioral intention toward drug use prevention, as well as facilitating the reduction of illicit drug use.
is a critical period for the onset of illicit drug use [2,3]. Compared with adult-onset drug users, adolescent-onset drug users experience a greater likelihood of dependence and more serious clinical syndromes [4,5]. Therefore, it is not surprising that intervention programs aiming to prevent the use of illicit drugs among adolescents are considered to be of paramount importance [6]. On the basis of the data from a national campus survey conducted in Taiwan in 2008e2012, the rates of illicit drug use
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
among junior high school students aged 13e15 years were relatively low, ranging from .48% to 1.04% [7]. However, the prevalence of male illicit drug users aged 16e18 years (1.5%) was double that of illicit drug users aged 13e15 years (average, .7%) [8]. Consequently, the issue of illicit drug use among high school students in Taiwan cannot be ignored, and we reasoned that an intervention directed toward discouraging drug use among students in the early-onset group (13e15 years) would help to reduce drug use among the older students (16e18 years) in whom the onset is more typical. Unlike drug use policies in the United States, which adopt a public health perspective toward users [9,10], illicit drug use in Taiwan is considered, especially by parents and school teachers, to be very serious misconduct representative of criminal behavior, and it is not at all tolerated. The Juvenile Delinquency Act in Taiwan, which is a public policy document, cites illegal drug use as an example of criminal behavior. Students using illicit drugs are subjected to the related Protective Measures and Treatment program (derived from “Section 2: Execution of Protective Measures” of the Juvenile Delinquency Act in Taiwan), which emphasizes law enforcement intervention. Accordingly, young drug users will usually be withdrawn from school, which creates a series of social problems subsequently, and are associated with a high risk of continued illicit drug use. Many school-based intervention/education programs are able to help prevent occasional and more serious drug use among low- to high-risk adolescents in diverse school environments [11,12]. We recently reported the development of a novel, theoretically based drug use prevention program for use in schools in Taiwan [13]. The program was designed by integrating life skills training into the theory of planned behavior (TPB). On the basis of that theory, practical life skills include problem solving, coping with stressful situations, and development of social and communication skills, which have broader applications than just preventing drug use [14]. Results of another drug use intervention program showed that after students increased their knowledge about life skills (e.g., communication, problem solving), their attitudes toward substance abuse (e.g., tobacco, alcohol, and nicotine use) became more negative [15]. Life skills training actually taught students the skills needed to resist social pressures to use illicit drugs and helped students develop personal self-management and social skills [16]. In Taiwan, it is especially important to include life skills in drug use prevention programs because most teachers in Taiwan are still using the teacher-centered traditional didactic teaching pedagogy. Except for teaching the skill of avoidance, general life skills to prevent drug use are missing in schoolbased education in Taiwan. We theorized that integrating life skills into a TPB-based drug use prevention program would compensate for the limitations of TPB and improve program efficacy. Therefore, this study integrated TPB and life skills into the design of an adolescent drug use prevention program, relying on the TPB to provide structural psychosocial measures for predicting behavioral intention and behavior, and applying life skills as a strategy for competency enhancement and behavior modification. The present longitudinal, cluster-randomized controlled trial aimed mainly to examine the effectiveness of our intervention program for preventing illicit drug use and to validate our previous findings [13]. We also examined participants’ attitudes, subjective norms, perceived behavioral control, life skills, and intentions regarding illicit drugs.
315
Methods Study design and participants We conducted a cluster-randomized trial in which participating schools were randomized to either an intervention group or a control group. The inclusion criteria for school selection were as follows: (1) school authorities agreed to fully participate in the main session and two booster sessions of the drug use prevention program; (2) the school had licensed health education teachers who were willing to participate in the program; and (3) the same health education teacher was able to perform both the main session and two booster sessions consistently throughout the program. A total of 268 junior high schools in Northern Taiwan were invited to participate, of which 24 met the inclusion criteria (24/268 ¼ 8.96%), agreed to participate, and were included in the study. The schools were then randomly assigned to either the intervention (12 schools) or control group (12 schools). All classes of grade 7 students in the respective schools were invited to take part in the study. A total of 94 grade 7 classes with a total of 2,091 students were finally enrolled to participate in the study, including 55 classes in the intervention group (n ¼ 1,176) and 39 classes in control group (n ¼ 915). Figure 1 summarizes the flow of participating schools and students in the study. A total of 10 schools (845 students) in the intervention group and 11 schools (830 students) in the control group completed the study. The participant retention rates were 71.9% (845/1,176) in the intervention group and 90.7% (830/915) in the control group. Ethical considerations The study protocol was approved by the institutional review board of National Yang-Ming University. Informed consent was obtained from all participating students and their parents/guardians. The students each completed the questionnaire anonymously; and research staff were blinded to the students’ group status. Drug use prevention program The contents and learning activities of the drug use prevention program were described in detail in our previous report [13]. The TPB/life skills school-based intervention program was designed by our research team by following the seven evidence-based quality criteria for developing an intervention program for drug use prevention [16]. The recommended evidence-based quality criteria for drug use prevention programs are as follows: (1) an interactive delivery method; (2) utilization of social influence; (3) a focus on social norms; (4) commitment not to use illicit drugs; (5) community interventions; (6) use of peer leadership; and (7) inclusion of life skills in the programs. The program contents of life skills training were developed on the basis of the definitions of life skills training proposed by the World Health Organization (WHO) Department of Mental Health (http://www.who.int/mental_ health/media/en/30.pdf), which consists of five domains such as: self-awareness and empathy, creative and critical thinking, coping with stress and emotions, decision-making and problem solving, and communication and interpersonal relationships. We carefully followed these criteria in developing the present program, and the drug use prevention education was carefully integrated into the routine heath education curriculum. In addition, two additional booster sessions were added after the main program was completed. Please refer to the “Intervention and procedures” section for details.
316
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Figure 1. Study flow diagram.
Teacher training session and the fidelity of program implementation Health education teachers who conducted the intervention received two half-days of training before the beginning of the program. The training session included detailed descriptions
of the rationale for the program, procedure/protocol for conducting the program, and the method of data collection. At the end of training session, the health education teachers received a package containing learning sheets and the manual of drug use prevention program for teachers and students. Regular meetings were scheduled by the principal investigators and
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
staff to monitor and supervise the teachers’ performance when they conducted the program. To monitor the fidelity of program implementation, a 14-item checklist was developed on the basis of the five domains (adherence, dose, quality of delivery, responsiveness, and program differentiation) suggested by Dane and Schneider (1998) [17]. The average score >85% (12/14) indicated that the program was feasible and reliable.
317
exercises, whereas the second booster comprised two 45-minute sessions and homework exercises. The booster sessions focused on reviewing the contents of the main program and practicing life skills that discourage the use of illicit drugs. Participants in the control group received conventional didactic teaching about the harmful effects of using illicit drugs and drug refusal skills for two 45-minute sessions. Outcome measures
Intervention and procedures The intervention included the main program and two booster sessions of the school-based drug use prevention program (Figure 1). The main program was designed to integrate life skills into the TPB structure for the seventh grade students. The program comprised 10 45-minute sessions delivered over 16 weeks and homework exercises to be completed in the following 8e10 weeks during the summer vacation. The two booster interventions were performed at 6 and 12 months after completion of the main program to help sustain program effectiveness, as demonstrated previously [18,19]. Table 1 summarizes the contents and learning activities of the booster sessions. The program contents and learning activities of the booster interventions were designed on the basis of the results of the pilot study we published previously [13], and furthermore modified on the basis of the outcome of in-depth interviews with selected participants of the pilot study 3 months after it was completed. The first booster comprised four 45-minute sessions and homework
Assessments were made using a structured self-reported questionnaire that participants were required to complete at four different time points during the program as follows: (1) at baseline (within 1 week before the intervention; T1); (2) within 1 week after the main intervention and homework exercises (T2); (3) within 1 week after the first booster session (T3); and (4) within 1 week after the second booster session (T4) (Figure 1). The questionnaire was modified from a questionnaire described in our previous study [13], including turning certain items with negative questioning into positive questioning. The questionnaire underwent an expert review by a panel of five experts in drug use prevention before we conducted a pilot study, and the content validity index (CVI) was .95 (CVI was calculated for each item by dividing the number of experts who rated the item above three points by the total number of experts; the overall CVI was estimated by calculating the average CVI across items, as previously described [20]). Before beginning the program, students received instruction about the nature of the questionnaire and
Table 1 Overview of booster sessions for the school-based drug use prevention program for adolescents
Attitudes
Subjective norms
Perceived behavioral control
Life skills
Behavioral intention
Program contents
Learning activities
Build up positive values for rejecting illicit drugs (booster 1 and 2) Negative consequences of illicit drug use (boosters 1 and 2) Advance assertiveness and self-realization (booster 1) Personal insight and self-awareness of drug use (booster 2) Assertiveness skills (booster 2) Value clarification in difficult situations or in a dilemma (booster 2) Highlight the fact that most peers do not use illicit drugs and acceptability among adolescents (boosters 1 and 2) Correct erroneous perceptions of pressure and influence from others (booster 1) Testimonies from former drug users, with an emphasis on the social consequences of drug use (booster 2)
Questions and answers (booster 1 and 2) Animated film followed by group discussions (booster 1 and 2) Teacher paired with peer leader to address expected values (boosters 1 and 2) Completion of worksheets (booster 1)
Reinforce the skills of protecting self and peers in high-risk situations for drug use (boosters 1 and 2) Practice drug-related refusal skills (boosters 1 and 2) Assistance given to at risk individuals (boosters 1 and 2) Application of rational procedures to deal with problematic situations (booster 1) Personal commitment not to use drugs (booster 2) Creative and critical thinking (boosters 1 and 2) Decision-making and problem solving (boosters 1 and 2) Empathy (booster 1) Self-management (booster 1) Communication and interpersonal relationships (booster 1) Assertiveness skills and refusal skills (booster 1) Goal setting (booster 2) Self-awareness (booster 2) Action plan (boosters 1 and 2) Commitment (boosters 1 and 2) Provide monitoring to each other (boosters 1 and 2)
Animated film followed by group discussions (boosters 1 and 2) Provide prompt/credible feedback (boosters 1 and 2) Discuss acceptability of group norms in peer-oriented social setting (booster 1) Use of peer leader in discussion (booster 1) Completion of worksheets (booster 2) Group discussion (boosters 1 and 2) Situational role play (boosters 1 and 2) Give verbal reinforcement and prizes (boosters 1 and 2) Parents/guardian involvement (homework exercise) (boosters 1 and 2) Instructional guidance on skill concept formation (booster 1) Information and arrangement about resources (booster 2) Modeling and skill practices (boosters 1 and 2) Situational role playing to demonstrate skills (boosters 1 and 2) Homework exercises on skill maintenance (boosters 1 and 2) Experience sharing (booster 1)
Sign up the worksheet (boosters 1 and 2) Public demonstration (boosters 1 and 2) Behavior log and exchange, provide feedback (boosters 1 and 2)
318
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
how it would be scored using the seven-point Likert-type scale. The items in the questionnaire addressed the following: Part I, Demographic and substance abuse information, including demographic information (age, sex, educational level and occupation of the main guardian, household status, and religion) and substance abuse (illicit drug use defined as use of amphetamines, ketamines, or Ecstasy), illicit drug use of family members and friends, cigarette smoking, and alcohol consumption; Part II, TPB variables and life skills, including four TPB variables (attitudes, subjective norms, perceived behavioral control, and behavioral intention) and life skills to prevent using illicit drugs and substance use behaviors. The TPB measures were designed according to the TPB manual for constructing questionnaires [21], and life skills were measured by questionnaires designed by our research team on the basis of the definitions of life skills proposed by the WHO Department of Mental Health (http://www. who.int/mental_health/media/en/30.pdf). Illicit drug use, smoking, and drinking. In the self-reported questionnaire (Part I), participants would check “Yes” if they had used illicit drugs, smoked cigarettes, or drank alcohol. “Illicit drug use” refers to the use of illegal drugs such as amphetamine, ketamine, Ecstasy, and/or other illegal drugs. All measures of drug use refer to present usage and usage within the previous month. Theory of planned behavior variables. For the assessment of attitudes, participants were asked to rate their level of feelings about happiness or interest associated with illicit drug use, such as “It is not a joyful experience to use an illicit drug.” Scores for the fouritem attitudes scale ranged from 4 to 28 using a seven-point Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). A higher score indicated that a more positive attitude toward not using illicit drugs. Cronbach’s a was .90e.98 across the four measurements at different times during the program. For the assessment of subjective norms, participants were asked if their parents, teachers, classmates, friends, and other people expected them to refuse illicit drugs, such as “My parents/ guardians think I should not use illicit drugs.” Scores for the fiveitem subjective norms scale ranged from 5 to 35 using a sevenpoint Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). A higher score indicated that the students strongly perceived a disapproving subjective norm toward drug use. Cronbach’s a was .85e.96 across the four measurements at different times during the program. For the assessment of perceived behavioral control, participants were asked to rate their ability to avoid illicit drug use, such as “Whether to use drugs or not is up to me.” Scores for the threeitem perceived behavioral control scale ranged from 3 to 21 using a seven-point Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicated greater control over not using drugs. Cronbach’s a was .90e.97 across the four measurements at different times during the program. For the assessment of behavioral intention, participants were asked to rate the statements, such as “I don’t intend to use illicit drugs,” from 1 (strongly disagree) to 7 (strongly agree). Participants were also asked to report “Given 10 occasions in which you have the chance to use illegal drugs, how many times would you expect not to use them?” The Likert-type scoring ranged from 0 (10 times) to 10 (refuse to use drugs consistently) for this item was used. Thus, the scores for the three-item behavioral intervention scale ranged from 2 to 24. Higher scores indicated greater intention toward not using drugs. Cronbach’s a was
.80e.92 across the four measurements at different times during the program. Life skills directed toward not using illicit drugs. To assess life skills directed toward not using illicit drugs, participants were asked to consider 16 hypothetical situations on the basis of the definitions proposed by the WHO Department of Mental Health (http:// www.who.int/mental_health/media/en/30.pdf). A question for creative and critical thinking was “I can judge the advantages and disadvantages of drug use by their health effects and social consequences.” A question for coping with stress and emotions was “When feeling upset or bored, I won’t use any drug to change my mood.” A question for communication and interpersonal relationships was “I can effectively refuse to use drugs when someone offers me drugs.” A question for self-awareness and empathy was “I know what my reactions or behavior may be when I am in a bad mood (e.g., depressed or irritable).” A question for decision-making and problem solving was “Before deciding whether to use drugs or not, I can think over the possible consequences.” Scores for the 16-item scale ranged from 16 to 112 using a seven-point Likert-type scoring from 1 (extreme unlikely) to 7 (extreme likely). A higher score indicated better performance in life skills toward not using drugs. Cronbach’s a was .93e.97 across the four measurements at different times during the program. Statistical analysis Categorical variables were presented as count and percentage. Continuous variables were normally distributed and presented as mean and standard deviation. Differences between the intervention group and the control group were compared by an independent two-sample t test (continuous variables) or the chisquare test/Fisher exact test with Yates correction (categorical variables). Because of the repeated measurements of outcome variables over time, a linear mixed model with the first-order autoregressive [AR(1)] structure was used to investigate the effects of time (denoted as Time effect), groups (denoted as Group effect), and the effects of their interactions (denoted as Group Time interaction), respectively. When main effects or interaction showed significance, further post hoc multiple comparisons were conducted using a Bonferroni correction to control for overall type I error rates. All statistical assessments were twosided and evaluated at the .05 level of statistical significance. Statistical analyses were performed using SPSS 18.0 (SPSS Inc., Chicago, IL). Because illicit drug use is binary (Yes vs. No), no clustering effect was applied to measurements. “Behavioral intention” was considered as an independent variable, and the intraclass correlation coefficient (ICC) was analyzed using HLM6 (SSI, Inc., Skokie, IL). On the basis of interschool variations (s00) and intraschool variations (s2) calculated from the random effects analysis of variance model, the ICC (r) ＝ s00/(s00 ＋ s2) ＝ .039/ (.039 ＋ 3.593) ＝ .011. On the basis of the standard proposed by Cohen (1988) [22], ICC < .059 can be ignored. This suggests that the interschool variation only contributes 1.1% of the overall variation, and that the major variation of “behavioral intention” measurement (98.9%) came from intraschool variations. Sensitivity analysis and attriter-remainder comparison analysis were performed to evaluate whether effects of differential attrition across conditions during the study period influenced the study outcomes.
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Results The demographic characteristics of the two study groups are summarized in Table 2. The two groups were similar in age (13.39 .55 years and 13.43 .60, respectively) and gender (male students, 51.1% and 54.1%, respectively). Moreover, no significant differences were observed between the two groups in other baseline variables (Table 2). The number of dropouts is shown in Figure 1. Table 3 shows the proportion of participants reporting illicit drug use, cigarette smoking, and consumption of alcohol at baseline (T1), after the main intervention (T2), after the first booster session (T3), and after the second booster session (T4). No between-group differences were found in the proportion of participants reporting illicit drug use at T1 or T2, whereas a significantly lower proportion of participants in the intervention group reported illicit drug use at T3 and T4 compared with the control group (both p < .05). No significant between-group differences were found in the proportion of participants reporting cigarette smoking or consumption of alcohol at any time point (Table 3). Figure 2 summarizes the changes in attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention scores for the two groups. Linear mixed models revealed significant differences in terms of Group effect and Group Time interaction for all variables (all p < .001). The model also showed significant differences in Time effect for subjective norms, perceived behavioral control, life skills, and behavioral intention (all p < .001). Scores for attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention were significantly higher in the intervention group
Table 2 Demographic characteristics of participants in the intervention and comparison groups at baseline Characteristic
Intervention (n ¼ 1,176)
Control (n ¼ 915)
p value
Age (years) Gender Male Female Work status of main guardian Employed Unemployed Years of education of main guardiana <9 years 9 years Household status Live with both parents Live with single parent Live without parent Religion Yes No Illicit drug use among family members Yes No Illicit drug use among friends Yes No
13.39 .55
13.43 .60
.093 .174
601 (51.1) 575 (48.9)
495 (54.1) 420 (45.9)
compared with those in the control group at T3 and T4 (all p < .001). Subjective norms scores were also significantly higher in the intervention group compared with the control group at T2 (p ¼ .027), whereas life skills scores were significantly higher in the control group at T1 (p < .001) (Figure 2). Results of sensitivity analysis and attriter-remainder comparison analysis are shown in Supplementary Tables 1 and 2, respectively. A total of 716 pairs of dropout-matched cases and controls were found in the original analysis data. The results of sensitivity analysis showed similar patterns between original and dropout-matched data except for the result of subjective norm. However, results after the first and second booster courses (T3, T4) were consistent (Supplementary Table 1). Results indicate the robustness of findings after evaluation of differential attrition. For the purpose of attriter-remainder comparison analysis, attriters and remainders were identified by their participant status in specific time periods (i.e., T2, T3, T4) and main variables were compared with those of their prior performance (i.e., T1, T2, T3). Results revealed that no significant differences existed between attriters and remainders across variables of major outcomes, indicating that the potential for bias was minimal (Supplementary Table 2). Discussion In this study, we used a longitudinal, cluster-randomized design to examine the effects of a school-based drug use prevention program that integrated aspects of the TPB and life skills. To our knowledge, no other intervention programs designed to prevent illicit drug use in adolescents have incorporated both TPB and life skills. The intervention program, comprising a main
Table 3 Proportion of participants reporting illicit drug use, cigarette smoking, consumption of alcohol, and betel nut chewing at baseline (T1), after the main intervention (T2), after the first booster session (T3), and after the second booster session (T4)
Illicit drug usec .056
1,129 (96.0) 47 (4.0)
862 (94.2) 53 (5.8)
266 (22.9) 897 (77.1)
191 (21.2) 710 (78.8)
1,004 (85.4) 95 (8.1) 77 (6.5)
769 (84.0) 95 (10.4) 51 (5.6)
595 (50.6) 581 (49.4)
476 (52.0) 439 (48.0)
.364
Smokingd
.141
Alcohole
.517
.230 52 (4.4) 1,124 (95.6)
31 (3.4) 884 (96.6)
79 (6.7) 1,097 (93.3)
64 (7.0) 851 (93.0)
.804
Data are presented as mean standard deviation or number (percent) and were compared by an independent two-sample t test or the chi-square test. a There were 27 missing data points for this variable, 13 in the intervention group and 14 in the control group.
319
Time
Interventiona
Controlb
p value
T1 T2 T3 T4 T1 T2 T3 T4 T1 T2 T3 T4
3 5 1 2 33 28 28 33 30 36 34 40
8 8 14 14 23 25 27 30 28 33 28 40
.068 .184 <.001* .002* .679 .632 .834 .754 .484 .509 .903 .935
(.3) (.4) (.1) (.2) (2.8) (2.4) (3.1) (3.9) (2.6) (3.1) (3.4) (4.7)
(.9) (.9) (1.7) (1.7) (2.5) (2.8) (3.3) (3.6) (3.1) (3.7) (3.3) (4.8)
Data are expressed as number (percent) of participants in each group and were compared by the chi-square test or Fisher exact test. a Number of participants in intervention group are as follows: T1, 1,176; T2, 1,153; T3, 987; T4, 845. b Number of participants in control group are as follows: T1, 915; T2, 904; T3, 838; T4, 830. c There were 22 missing data points for this variable at T2, all in the control group. d There was one missing data point for this variable at T1 in the intervention group and 85 missing data points at T3 (77 in the intervention group and 8 in the control group). e There was one missing data point for this variable at T1 in the intervention group. *p < .05 indicates a significant difference between the two groups.
320
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Figure 2. Changes in the theory of planned behavior variables and life skills for participants in the intervention and control groups include the results of the linear mixed model. (A) ATT ¼ attitudes. (B) SN ¼ subjective norms. (C) PBC ¼ perceived behavioral control. (D) LS ¼ life skills. (E) BI ¼ behavioral intention not to use illicit drugs. T1 ¼ baseline; T2 ¼ after the main intervention program; T3 ¼ after the first booster session; T4 ¼ after the second booster session. *Significant difference (p < .05) between the two groups. Error bar indicates standard deviation.
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
intervention and two booster sessions, was implemented at 12 middle schools among students aged 13e15 years, and showed evidence of effectiveness as indicated by the analysis of findings from students’ self-reported questionnaires assessing illicit drug use and changes in TPB indicators and life skills from the beginning of the program until 12 months after main intervention completion. In our previously published pilot study of the TPB/life skills school-based intervention program [13], we found that participants who received the intervention (in 10 45-minute sessions) demonstrated superior performance in attitudes, subjective norms, perceived behavioral control, and life skills compared with participants in the control/nonintervention group. Furthermore, participants in the intervention group were also found to have significantly greater behavioral intention not to use illicit drugs than participants in the control group [13]. Although the present study was designed on the basis of the pilot study, some aspects of the present implementation were unique. First, the illicit drug prevention program was modified by adding two booster sessions at 6 and 12 months after the second assessment. Second, the two booster sessions allowed us to collect follow-up data at 6 months (T3) and 12 months (T4), partially representing the long-term effect of the intervention program. From the beginning of the program to the second booster session, the students were progressing from the eighth grade to the ninth grade. Third, the booster session indeed provided an opportunity to reinforce the topic(s), especially the topics that students had not mastered in the previous assessment. White and Pitts (1997) [23] suggested that booster sessions provide the opportunity to reinforce and build on messages over a number of years suited to the age and development of the students. It is notable that the prevalence of illicit drug use at the end of program was increased in the control group (from .9% at T1 to 1.7% at T4) but remained unchanged in the intervention group (from .3% at T1 to .2% at T4). This suggests that our illicit drug use prevention program has the potential to effectively prevent (prophylactic effect) high school students from using illicit drugs in adulthood. Consistent with this finding, participants in the intervention group also had significantly higher scores for TPB and life skill measures at the end of the booster sessions. These findings indicate that the incorporation of TPB and life skills into the intervention program helped promote participants’ attitudes and subjective norms, and may potentially facilitate reducing illicit drug use. Our findings are in keeping with those from previous TPB studies that have demonstrated associations between attitudes and intentions and consequent illicit drug (marijuana) use [24] and between attitudes and subjective norms and the intention to use illicit drugs [25]. The effectiveness of skills-based drug use prevention programs has been demonstrated by a systematic review. A systematic review of 29 randomized control trials on drug abuse prevention programs, of which the majority were conducted in the United States among the sixth and seventh graders just slightly younger than our target groups, found that skills-based interventions significantly reduced marijuana use and hard drug use, and improved decision-making skills, self-esteem, and peer pressure resistance compared with conventional curricula [16]. A key part of our intervention program was the integration of two booster sessions that were implemented at 6 and 12 months after completing the initial intervention. It is interesting that, for all domains tested in the questionnaire, the scores between
321
groups only showed significant differences at the end of each booster session but not at the end of the 10-session main program (the second assessment; T2). This observation may result from two possible reasons: (1) the second assessment (T2) was not performed at the end of the 10-session main program but at the end of the 8- to 10-week summer vacation. Students were given homework exercises to complete during their summer vacation, but they apparently did not take the homework seriously. Therefore, the intervention protocol may require revision to ask the students to complete the homework exercise immediately after the end of the 10-session program and (2) the lack of between-group differences immediately after the main intervention may simply be a time-related effect. For example, a previous study examining the effects of an online intervention program (without booster sessions) on subsequent drug use among girls in grades 7e9, found a significant difference in the rate of drug use between the intervention and control groups at 6 months after program completion but not immediately after the program [26]. The present study attempted to reduce possible biases and other confounders (e.g., cultural norms, expectations) by using an anonymous questionnaire. Students were informed before filling out the questionnaires that the outcome of the study would be for research use only, and only the research team in this program would be able to access student data. In addition, not all schools and students enrolled completed the study. As a consequence, the withdrawal of schools and students may have biased our findings to some extent. To rule out differential attrition as a threat to conclusion validity, we have conducted sensitivity analysis and attriter-remainder comparison analysis to confirm that differential attrition between the intervention and control groups in our study did not bias the study outcomes. Results of sensitivity analysis and comparison of attriter-remainder differences at different time points indicated that no significant differences existed between attriters and remainders in main outcome variables, confirming that the potential for bias due to differential attrition was minimal. Our study has a number of limitations that warrant mention. First, we did not compare the effectiveness of our intervention with other evidence-based drug prevention programs for middle school students such as Project ALERT in the United States [9] or the LifeSkills Training Program outlined by Botvin and Kantor [15]. Given that differences exist in cultural norms and practices, the program implemented in the present study may not be readily applicable to other populations and vice versa. Second, we cannot comment on the longer term effectiveness (i.e., a number of years later when the students become adults) of our intervention program. Additional follow-up is needed to determine how the described intervention program affects illicit drug use beyond 12 months after completion of the main intervention. In conclusion, after participants completed the main intervention and two booster sessions of the theoretically based drug use prevention program, the scores for attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention were significantly higher among students receiving the intervention than among the control groups receiving conventional antidrug use didactic teaching. These results suggest that our drug use prevention program that integrates life skills into the TPB may be effective for reducing illicit drug use and improving planned behavior-related constructs in adolescents. Pending further research and confirmation of the findings reported herein, practical implementation of this program may be warranted.
322
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Acknowledgments The authors thank all students and teachers who participated in the study. Funding Sources This work was supported by grants from the Taiwan Food and Drug Administration, Department of Health (DOH 99-FDA61209; DOH 100-FDA-61202). This article was subsidized by the National Taiwan Normal University, Taiwan, ROC. Supplementary Data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jadohealth.2014.12.003. References [1] Degenhardt L, Hall W. Extent of illicit drug use and dependence, and their contribution to the global burden of disease. Lancet 2012;379:55e70. [2] Swendsen J, Burstein M, Case B, et al. Use and abuse of alcohol and illicit drugs in U.S. adolescents: Results of the National Comorbidity SurveyAdolescent Supplement. Arch Gen Psychiatry 2012;69:390e8. [3] Vega WA, Aguilar-Gaxiola S, Andrade L, et al. Prevalence and age of onset for drug use in seven international sites: Results from the international consortium of psychiatric epidemiology. Drug Alcohol Depend 2002;68:285e97. [4] Chen CY, Storr CL, Anthony JC. Early-onset drug use and risk for drug dependence problems. Addict Behav 2009;34:319e22. [5] Rubino T, Zamberletti E, Parolaro D. Adolescent exposure to cannabis as a risk factor for psychiatric disorders. J Psychopharmacol 2012;26:177e88. [6] Botvin GJ, Griffin KW. School-based programmes to prevent alcohol, tobacco and other drug use. Int Rev Psychiatry 2007;19:607e15. [7] Guo JL, Huang CM, Lee TC. National Web-based survey of illegal drug use among adolescents. Taipei, Taiwan: Ministry of Education; 2008. [8] Chen KT, Chen CJ, Fagot-Campagna A, et al. Tobacco, betel quid, alcohol, and illicit drug use among 13- to 35-year-olds in I-Lan, rural Taiwan: Prevalence and risk factors. Am J Public Health 2001;91:1130e4. [9] McBride DC, Terry-McElrath YM, VanderWaal CJ, et al. U.S. Public Health Agency involvement in youth-focused illicit drug policy, planning, and prevention at the local level, 1999-2003. Am J Public Health 2008;98:270e7.
[10] Des Jarlais DC. Prospects for a public health perspective on psychoactive drug use. Am J Public Health 2000;90:335e7. [11] Ellickson PL, McCaffrey DF, Ghosh-Dastidar B, et al. New inroads in preventing adolescent drug use: Results from a large-scale trial of project ALERT in middle schools. Am J Public Health 2003;93: 1830e6. [12] Griffin KW, Botvin GJ. Evidence-based interventions for preventing substance use disorders in adolescents. Child Adolesc Psychiatr Clin N Am 2010;19:505e26. [13] Huang CM, Chien LY, Cheng CF, et al. Integrating life skills into a theorybased drug-use prevention program: Effectiveness among junior high students in Taiwan. J Sch Health 2012;82:328e35. [14] Buhler A, Schroder E, Silbereisen RK. The role of life skills promotion in substance abuse prevention: A mediation analysis. Health Educ Res 2008; 23:621e32. [15] Botvin GJ, Kantor LW. Preventing alcohol and tobacco use through life skills training. Alcohol Res Health 2000;24:250e7. [16] Faggiano F, Vigna-Taglianti FD, Versino E, et al. School-based prevention for illicit drugs use: A systematic review. Prev Med 2008;46:385e96. [17] Dane AV, Schneider BH. Program integrity in primary and early secondary prevention: Are implementation effects out of control? Clin Psychol Rev 1998;18:23e45. [18] Botvin GJ. Preventing drug abuse in schools: Social and competence enhancement approaches targeting individual-level etiologic factors. Addict Behav 2000;25:887e97. [19] McBride N. A systematic review of school drug education. Health Educ Res 2003;18:729e42. [20] Rubio DM, Berg-Weber M, Tebb SS, et al. Objectifying content validity: Conducting a content validity study in social work research. Soc Work Res 2003;27:94e104. [21] Francis JJ, Eccles MP, Johnston M, et al. Constructing questionnaires based on the theory of planned behaviour: A manual for health services researchers. University of Newcastle upon Tyne: Centre for Health Services Research, Newcastle upon Tyne; 2004. [22] Cohen, J. Statistical power analysis for the behavioral sciences (2nd ed.). Hillside, New Jersey: Lawrence Erlbaum Associates;1988. [23] White D, Pitts M. Educating young people about drugs: A systematic review. Addiction 1998;93:1475e87. [24] Hohman ZP, Crano WD, Siegel JT, et al. Attitudes ambivalence, friend norms, and adolescent drug use. Prev Sci 2014;15:65e74. [25] Bashirian S, Hidarnia A, Allahverdipour H, et al. Application of the theory of planned behavior to predict drug abuse related behaviors among adolescents. J Res Health Sci 2012;12:54e60. [26] Schwinn TM, Schinke SP, Di Noia J. Preventing drug abuse among adolescent girls: Outcome data from an Internet-based intervention. Prev Sci 2010;11:24e32.
www.jahonline.org Original article
Prevention of Illicit Drug Use Through a School-Based Program: Results of a Longitudinal, Cluster-Randomized Controlled Trial Jong-Long Guo, Ph.D. a, Tzu-Chi Lee, Ph.D. b, Jung-Yu Liao, M.Ed. a, and Chiu-Mieh Huang, Ph.D. c, * a b c
Department of Health Promotion and Health Education, National Taiwan Normal University, Taipei, Taiwan Department of Public Health, Kaohsiung Medical University, Kaohsiung, Taiwan School of Nursing, National Yang-Ming University, Taipei, Taiwan
Article history: Received October 16, 2013; Accepted December 1, 2014 Keywords: Illicit drug use prevention program; Booster session; Theory of planned behavior; Life skills; Perceived behavioral control; Behavioral intention
A B S T R A C T
Purpose: To evaluate the long-term effects of an illicit drug use prevention program for adolescents that integrates life skills into the theory of planned behavior. Methods: We conducted a cluster-randomized trial in which 24 participating schools were randomized to either an intervention group (12 schools, n ¼ 1,176 students) or a control group (12 schools, n ¼ 915 students). Participants were grade 7 students. The intervention comprised a main intervention of 10 sessions and two booster interventions. Booster 1 (four sessions) and booster 2 (two sessions) were performed at 6 months and 12 months, respectively, after completion of the main intervention. Assessments were made at baseline, after the main intervention, and after each booster session using specific questionnaires for measuring participants’ attitudes, subjective norms, perceived behavioral control, and life skills. Results: Retention rates were 71.9% (845/1,176) in the intervention group and 90.7% (830/915) in the control group after the 12-month follow-up. A significantly lower proportion of intervention group participants reported illicit drug use after the first and second booster sessions compared with control group participants (.1% vs. 1.7% and .2% vs. 1.7%, respectively; both p < .05). Attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention scores of the intervention group were significantly higher than those of control group after the first and second booster sessions (all p < .001), suggesting that intervention group students tended to avoid drug use. Conclusions: A drug use prevention program integrating life skills into the theory of planned behavior may be effective for reducing illicit drug use and improving planned behavior-related constructs in adolescents. Ó 2015 Society for Adolescent Health and Medicine. All rights reserved.
Illicit drug use can have a number of negative consequences, including physical and mental disorders, unemployment, accidents, suicide, violence, and a significant financial burden [1]. Adolescence Conflicts of Interest: The authors have no conflicts of interest to declare. * Address correspondence to: Chiu-Mieh Huang, Ph.D., School of Nursing, National Yang-Ming University, 155, Linong Street, Section 2, Taipei, Taiwan 11221. E-mail address: [email protected] (C.-M. Huang). 1054-139X/Ó 2015 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/j.jadohealth.2014.12.003
IMPLICATIONS AND CONTRIBUTION
An intervention that incorporated the theory of planned behavior and life skills helped to promote participants’ attitudes, subjective norms, perceived behavioral control, and behavioral intention toward drug use prevention, as well as facilitating the reduction of illicit drug use.
is a critical period for the onset of illicit drug use [2,3]. Compared with adult-onset drug users, adolescent-onset drug users experience a greater likelihood of dependence and more serious clinical syndromes [4,5]. Therefore, it is not surprising that intervention programs aiming to prevent the use of illicit drugs among adolescents are considered to be of paramount importance [6]. On the basis of the data from a national campus survey conducted in Taiwan in 2008e2012, the rates of illicit drug use
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
among junior high school students aged 13e15 years were relatively low, ranging from .48% to 1.04% [7]. However, the prevalence of male illicit drug users aged 16e18 years (1.5%) was double that of illicit drug users aged 13e15 years (average, .7%) [8]. Consequently, the issue of illicit drug use among high school students in Taiwan cannot be ignored, and we reasoned that an intervention directed toward discouraging drug use among students in the early-onset group (13e15 years) would help to reduce drug use among the older students (16e18 years) in whom the onset is more typical. Unlike drug use policies in the United States, which adopt a public health perspective toward users [9,10], illicit drug use in Taiwan is considered, especially by parents and school teachers, to be very serious misconduct representative of criminal behavior, and it is not at all tolerated. The Juvenile Delinquency Act in Taiwan, which is a public policy document, cites illegal drug use as an example of criminal behavior. Students using illicit drugs are subjected to the related Protective Measures and Treatment program (derived from “Section 2: Execution of Protective Measures” of the Juvenile Delinquency Act in Taiwan), which emphasizes law enforcement intervention. Accordingly, young drug users will usually be withdrawn from school, which creates a series of social problems subsequently, and are associated with a high risk of continued illicit drug use. Many school-based intervention/education programs are able to help prevent occasional and more serious drug use among low- to high-risk adolescents in diverse school environments [11,12]. We recently reported the development of a novel, theoretically based drug use prevention program for use in schools in Taiwan [13]. The program was designed by integrating life skills training into the theory of planned behavior (TPB). On the basis of that theory, practical life skills include problem solving, coping with stressful situations, and development of social and communication skills, which have broader applications than just preventing drug use [14]. Results of another drug use intervention program showed that after students increased their knowledge about life skills (e.g., communication, problem solving), their attitudes toward substance abuse (e.g., tobacco, alcohol, and nicotine use) became more negative [15]. Life skills training actually taught students the skills needed to resist social pressures to use illicit drugs and helped students develop personal self-management and social skills [16]. In Taiwan, it is especially important to include life skills in drug use prevention programs because most teachers in Taiwan are still using the teacher-centered traditional didactic teaching pedagogy. Except for teaching the skill of avoidance, general life skills to prevent drug use are missing in schoolbased education in Taiwan. We theorized that integrating life skills into a TPB-based drug use prevention program would compensate for the limitations of TPB and improve program efficacy. Therefore, this study integrated TPB and life skills into the design of an adolescent drug use prevention program, relying on the TPB to provide structural psychosocial measures for predicting behavioral intention and behavior, and applying life skills as a strategy for competency enhancement and behavior modification. The present longitudinal, cluster-randomized controlled trial aimed mainly to examine the effectiveness of our intervention program for preventing illicit drug use and to validate our previous findings [13]. We also examined participants’ attitudes, subjective norms, perceived behavioral control, life skills, and intentions regarding illicit drugs.
315
Methods Study design and participants We conducted a cluster-randomized trial in which participating schools were randomized to either an intervention group or a control group. The inclusion criteria for school selection were as follows: (1) school authorities agreed to fully participate in the main session and two booster sessions of the drug use prevention program; (2) the school had licensed health education teachers who were willing to participate in the program; and (3) the same health education teacher was able to perform both the main session and two booster sessions consistently throughout the program. A total of 268 junior high schools in Northern Taiwan were invited to participate, of which 24 met the inclusion criteria (24/268 ¼ 8.96%), agreed to participate, and were included in the study. The schools were then randomly assigned to either the intervention (12 schools) or control group (12 schools). All classes of grade 7 students in the respective schools were invited to take part in the study. A total of 94 grade 7 classes with a total of 2,091 students were finally enrolled to participate in the study, including 55 classes in the intervention group (n ¼ 1,176) and 39 classes in control group (n ¼ 915). Figure 1 summarizes the flow of participating schools and students in the study. A total of 10 schools (845 students) in the intervention group and 11 schools (830 students) in the control group completed the study. The participant retention rates were 71.9% (845/1,176) in the intervention group and 90.7% (830/915) in the control group. Ethical considerations The study protocol was approved by the institutional review board of National Yang-Ming University. Informed consent was obtained from all participating students and their parents/guardians. The students each completed the questionnaire anonymously; and research staff were blinded to the students’ group status. Drug use prevention program The contents and learning activities of the drug use prevention program were described in detail in our previous report [13]. The TPB/life skills school-based intervention program was designed by our research team by following the seven evidence-based quality criteria for developing an intervention program for drug use prevention [16]. The recommended evidence-based quality criteria for drug use prevention programs are as follows: (1) an interactive delivery method; (2) utilization of social influence; (3) a focus on social norms; (4) commitment not to use illicit drugs; (5) community interventions; (6) use of peer leadership; and (7) inclusion of life skills in the programs. The program contents of life skills training were developed on the basis of the definitions of life skills training proposed by the World Health Organization (WHO) Department of Mental Health (http://www.who.int/mental_ health/media/en/30.pdf), which consists of five domains such as: self-awareness and empathy, creative and critical thinking, coping with stress and emotions, decision-making and problem solving, and communication and interpersonal relationships. We carefully followed these criteria in developing the present program, and the drug use prevention education was carefully integrated into the routine heath education curriculum. In addition, two additional booster sessions were added after the main program was completed. Please refer to the “Intervention and procedures” section for details.
316
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Figure 1. Study flow diagram.
Teacher training session and the fidelity of program implementation Health education teachers who conducted the intervention received two half-days of training before the beginning of the program. The training session included detailed descriptions
of the rationale for the program, procedure/protocol for conducting the program, and the method of data collection. At the end of training session, the health education teachers received a package containing learning sheets and the manual of drug use prevention program for teachers and students. Regular meetings were scheduled by the principal investigators and
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
staff to monitor and supervise the teachers’ performance when they conducted the program. To monitor the fidelity of program implementation, a 14-item checklist was developed on the basis of the five domains (adherence, dose, quality of delivery, responsiveness, and program differentiation) suggested by Dane and Schneider (1998) [17]. The average score >85% (12/14) indicated that the program was feasible and reliable.
317
exercises, whereas the second booster comprised two 45-minute sessions and homework exercises. The booster sessions focused on reviewing the contents of the main program and practicing life skills that discourage the use of illicit drugs. Participants in the control group received conventional didactic teaching about the harmful effects of using illicit drugs and drug refusal skills for two 45-minute sessions. Outcome measures
Intervention and procedures The intervention included the main program and two booster sessions of the school-based drug use prevention program (Figure 1). The main program was designed to integrate life skills into the TPB structure for the seventh grade students. The program comprised 10 45-minute sessions delivered over 16 weeks and homework exercises to be completed in the following 8e10 weeks during the summer vacation. The two booster interventions were performed at 6 and 12 months after completion of the main program to help sustain program effectiveness, as demonstrated previously [18,19]. Table 1 summarizes the contents and learning activities of the booster sessions. The program contents and learning activities of the booster interventions were designed on the basis of the results of the pilot study we published previously [13], and furthermore modified on the basis of the outcome of in-depth interviews with selected participants of the pilot study 3 months after it was completed. The first booster comprised four 45-minute sessions and homework
Assessments were made using a structured self-reported questionnaire that participants were required to complete at four different time points during the program as follows: (1) at baseline (within 1 week before the intervention; T1); (2) within 1 week after the main intervention and homework exercises (T2); (3) within 1 week after the first booster session (T3); and (4) within 1 week after the second booster session (T4) (Figure 1). The questionnaire was modified from a questionnaire described in our previous study [13], including turning certain items with negative questioning into positive questioning. The questionnaire underwent an expert review by a panel of five experts in drug use prevention before we conducted a pilot study, and the content validity index (CVI) was .95 (CVI was calculated for each item by dividing the number of experts who rated the item above three points by the total number of experts; the overall CVI was estimated by calculating the average CVI across items, as previously described [20]). Before beginning the program, students received instruction about the nature of the questionnaire and
Table 1 Overview of booster sessions for the school-based drug use prevention program for adolescents
Attitudes
Subjective norms
Perceived behavioral control
Life skills
Behavioral intention
Program contents
Learning activities
Build up positive values for rejecting illicit drugs (booster 1 and 2) Negative consequences of illicit drug use (boosters 1 and 2) Advance assertiveness and self-realization (booster 1) Personal insight and self-awareness of drug use (booster 2) Assertiveness skills (booster 2) Value clarification in difficult situations or in a dilemma (booster 2) Highlight the fact that most peers do not use illicit drugs and acceptability among adolescents (boosters 1 and 2) Correct erroneous perceptions of pressure and influence from others (booster 1) Testimonies from former drug users, with an emphasis on the social consequences of drug use (booster 2)
Questions and answers (booster 1 and 2) Animated film followed by group discussions (booster 1 and 2) Teacher paired with peer leader to address expected values (boosters 1 and 2) Completion of worksheets (booster 1)
Reinforce the skills of protecting self and peers in high-risk situations for drug use (boosters 1 and 2) Practice drug-related refusal skills (boosters 1 and 2) Assistance given to at risk individuals (boosters 1 and 2) Application of rational procedures to deal with problematic situations (booster 1) Personal commitment not to use drugs (booster 2) Creative and critical thinking (boosters 1 and 2) Decision-making and problem solving (boosters 1 and 2) Empathy (booster 1) Self-management (booster 1) Communication and interpersonal relationships (booster 1) Assertiveness skills and refusal skills (booster 1) Goal setting (booster 2) Self-awareness (booster 2) Action plan (boosters 1 and 2) Commitment (boosters 1 and 2) Provide monitoring to each other (boosters 1 and 2)
Animated film followed by group discussions (boosters 1 and 2) Provide prompt/credible feedback (boosters 1 and 2) Discuss acceptability of group norms in peer-oriented social setting (booster 1) Use of peer leader in discussion (booster 1) Completion of worksheets (booster 2) Group discussion (boosters 1 and 2) Situational role play (boosters 1 and 2) Give verbal reinforcement and prizes (boosters 1 and 2) Parents/guardian involvement (homework exercise) (boosters 1 and 2) Instructional guidance on skill concept formation (booster 1) Information and arrangement about resources (booster 2) Modeling and skill practices (boosters 1 and 2) Situational role playing to demonstrate skills (boosters 1 and 2) Homework exercises on skill maintenance (boosters 1 and 2) Experience sharing (booster 1)
Sign up the worksheet (boosters 1 and 2) Public demonstration (boosters 1 and 2) Behavior log and exchange, provide feedback (boosters 1 and 2)
318
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
how it would be scored using the seven-point Likert-type scale. The items in the questionnaire addressed the following: Part I, Demographic and substance abuse information, including demographic information (age, sex, educational level and occupation of the main guardian, household status, and religion) and substance abuse (illicit drug use defined as use of amphetamines, ketamines, or Ecstasy), illicit drug use of family members and friends, cigarette smoking, and alcohol consumption; Part II, TPB variables and life skills, including four TPB variables (attitudes, subjective norms, perceived behavioral control, and behavioral intention) and life skills to prevent using illicit drugs and substance use behaviors. The TPB measures were designed according to the TPB manual for constructing questionnaires [21], and life skills were measured by questionnaires designed by our research team on the basis of the definitions of life skills proposed by the WHO Department of Mental Health (http://www. who.int/mental_health/media/en/30.pdf). Illicit drug use, smoking, and drinking. In the self-reported questionnaire (Part I), participants would check “Yes” if they had used illicit drugs, smoked cigarettes, or drank alcohol. “Illicit drug use” refers to the use of illegal drugs such as amphetamine, ketamine, Ecstasy, and/or other illegal drugs. All measures of drug use refer to present usage and usage within the previous month. Theory of planned behavior variables. For the assessment of attitudes, participants were asked to rate their level of feelings about happiness or interest associated with illicit drug use, such as “It is not a joyful experience to use an illicit drug.” Scores for the fouritem attitudes scale ranged from 4 to 28 using a seven-point Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). A higher score indicated that a more positive attitude toward not using illicit drugs. Cronbach’s a was .90e.98 across the four measurements at different times during the program. For the assessment of subjective norms, participants were asked if their parents, teachers, classmates, friends, and other people expected them to refuse illicit drugs, such as “My parents/ guardians think I should not use illicit drugs.” Scores for the fiveitem subjective norms scale ranged from 5 to 35 using a sevenpoint Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). A higher score indicated that the students strongly perceived a disapproving subjective norm toward drug use. Cronbach’s a was .85e.96 across the four measurements at different times during the program. For the assessment of perceived behavioral control, participants were asked to rate their ability to avoid illicit drug use, such as “Whether to use drugs or not is up to me.” Scores for the threeitem perceived behavioral control scale ranged from 3 to 21 using a seven-point Likert-type scoring from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicated greater control over not using drugs. Cronbach’s a was .90e.97 across the four measurements at different times during the program. For the assessment of behavioral intention, participants were asked to rate the statements, such as “I don’t intend to use illicit drugs,” from 1 (strongly disagree) to 7 (strongly agree). Participants were also asked to report “Given 10 occasions in which you have the chance to use illegal drugs, how many times would you expect not to use them?” The Likert-type scoring ranged from 0 (10 times) to 10 (refuse to use drugs consistently) for this item was used. Thus, the scores for the three-item behavioral intervention scale ranged from 2 to 24. Higher scores indicated greater intention toward not using drugs. Cronbach’s a was
.80e.92 across the four measurements at different times during the program. Life skills directed toward not using illicit drugs. To assess life skills directed toward not using illicit drugs, participants were asked to consider 16 hypothetical situations on the basis of the definitions proposed by the WHO Department of Mental Health (http:// www.who.int/mental_health/media/en/30.pdf). A question for creative and critical thinking was “I can judge the advantages and disadvantages of drug use by their health effects and social consequences.” A question for coping with stress and emotions was “When feeling upset or bored, I won’t use any drug to change my mood.” A question for communication and interpersonal relationships was “I can effectively refuse to use drugs when someone offers me drugs.” A question for self-awareness and empathy was “I know what my reactions or behavior may be when I am in a bad mood (e.g., depressed or irritable).” A question for decision-making and problem solving was “Before deciding whether to use drugs or not, I can think over the possible consequences.” Scores for the 16-item scale ranged from 16 to 112 using a seven-point Likert-type scoring from 1 (extreme unlikely) to 7 (extreme likely). A higher score indicated better performance in life skills toward not using drugs. Cronbach’s a was .93e.97 across the four measurements at different times during the program. Statistical analysis Categorical variables were presented as count and percentage. Continuous variables were normally distributed and presented as mean and standard deviation. Differences between the intervention group and the control group were compared by an independent two-sample t test (continuous variables) or the chisquare test/Fisher exact test with Yates correction (categorical variables). Because of the repeated measurements of outcome variables over time, a linear mixed model with the first-order autoregressive [AR(1)] structure was used to investigate the effects of time (denoted as Time effect), groups (denoted as Group effect), and the effects of their interactions (denoted as Group Time interaction), respectively. When main effects or interaction showed significance, further post hoc multiple comparisons were conducted using a Bonferroni correction to control for overall type I error rates. All statistical assessments were twosided and evaluated at the .05 level of statistical significance. Statistical analyses were performed using SPSS 18.0 (SPSS Inc., Chicago, IL). Because illicit drug use is binary (Yes vs. No), no clustering effect was applied to measurements. “Behavioral intention” was considered as an independent variable, and the intraclass correlation coefficient (ICC) was analyzed using HLM6 (SSI, Inc., Skokie, IL). On the basis of interschool variations (s00) and intraschool variations (s2) calculated from the random effects analysis of variance model, the ICC (r) ＝ s00/(s00 ＋ s2) ＝ .039/ (.039 ＋ 3.593) ＝ .011. On the basis of the standard proposed by Cohen (1988) [22], ICC < .059 can be ignored. This suggests that the interschool variation only contributes 1.1% of the overall variation, and that the major variation of “behavioral intention” measurement (98.9%) came from intraschool variations. Sensitivity analysis and attriter-remainder comparison analysis were performed to evaluate whether effects of differential attrition across conditions during the study period influenced the study outcomes.
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Results The demographic characteristics of the two study groups are summarized in Table 2. The two groups were similar in age (13.39 .55 years and 13.43 .60, respectively) and gender (male students, 51.1% and 54.1%, respectively). Moreover, no significant differences were observed between the two groups in other baseline variables (Table 2). The number of dropouts is shown in Figure 1. Table 3 shows the proportion of participants reporting illicit drug use, cigarette smoking, and consumption of alcohol at baseline (T1), after the main intervention (T2), after the first booster session (T3), and after the second booster session (T4). No between-group differences were found in the proportion of participants reporting illicit drug use at T1 or T2, whereas a significantly lower proportion of participants in the intervention group reported illicit drug use at T3 and T4 compared with the control group (both p < .05). No significant between-group differences were found in the proportion of participants reporting cigarette smoking or consumption of alcohol at any time point (Table 3). Figure 2 summarizes the changes in attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention scores for the two groups. Linear mixed models revealed significant differences in terms of Group effect and Group Time interaction for all variables (all p < .001). The model also showed significant differences in Time effect for subjective norms, perceived behavioral control, life skills, and behavioral intention (all p < .001). Scores for attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention were significantly higher in the intervention group
Table 2 Demographic characteristics of participants in the intervention and comparison groups at baseline Characteristic
Intervention (n ¼ 1,176)
Control (n ¼ 915)
p value
Age (years) Gender Male Female Work status of main guardian Employed Unemployed Years of education of main guardiana <9 years 9 years Household status Live with both parents Live with single parent Live without parent Religion Yes No Illicit drug use among family members Yes No Illicit drug use among friends Yes No
13.39 .55
13.43 .60
.093 .174
601 (51.1) 575 (48.9)
495 (54.1) 420 (45.9)
compared with those in the control group at T3 and T4 (all p < .001). Subjective norms scores were also significantly higher in the intervention group compared with the control group at T2 (p ¼ .027), whereas life skills scores were significantly higher in the control group at T1 (p < .001) (Figure 2). Results of sensitivity analysis and attriter-remainder comparison analysis are shown in Supplementary Tables 1 and 2, respectively. A total of 716 pairs of dropout-matched cases and controls were found in the original analysis data. The results of sensitivity analysis showed similar patterns between original and dropout-matched data except for the result of subjective norm. However, results after the first and second booster courses (T3, T4) were consistent (Supplementary Table 1). Results indicate the robustness of findings after evaluation of differential attrition. For the purpose of attriter-remainder comparison analysis, attriters and remainders were identified by their participant status in specific time periods (i.e., T2, T3, T4) and main variables were compared with those of their prior performance (i.e., T1, T2, T3). Results revealed that no significant differences existed between attriters and remainders across variables of major outcomes, indicating that the potential for bias was minimal (Supplementary Table 2). Discussion In this study, we used a longitudinal, cluster-randomized design to examine the effects of a school-based drug use prevention program that integrated aspects of the TPB and life skills. To our knowledge, no other intervention programs designed to prevent illicit drug use in adolescents have incorporated both TPB and life skills. The intervention program, comprising a main
Table 3 Proportion of participants reporting illicit drug use, cigarette smoking, consumption of alcohol, and betel nut chewing at baseline (T1), after the main intervention (T2), after the first booster session (T3), and after the second booster session (T4)
Illicit drug usec .056
1,129 (96.0) 47 (4.0)
862 (94.2) 53 (5.8)
266 (22.9) 897 (77.1)
191 (21.2) 710 (78.8)
1,004 (85.4) 95 (8.1) 77 (6.5)
769 (84.0) 95 (10.4) 51 (5.6)
595 (50.6) 581 (49.4)
476 (52.0) 439 (48.0)
.364
Smokingd
.141
Alcohole
.517
.230 52 (4.4) 1,124 (95.6)
31 (3.4) 884 (96.6)
79 (6.7) 1,097 (93.3)
64 (7.0) 851 (93.0)
.804
Data are presented as mean standard deviation or number (percent) and were compared by an independent two-sample t test or the chi-square test. a There were 27 missing data points for this variable, 13 in the intervention group and 14 in the control group.
319
Time
Interventiona
Controlb
p value
T1 T2 T3 T4 T1 T2 T3 T4 T1 T2 T3 T4
3 5 1 2 33 28 28 33 30 36 34 40
8 8 14 14 23 25 27 30 28 33 28 40
.068 .184 <.001* .002* .679 .632 .834 .754 .484 .509 .903 .935
(.3) (.4) (.1) (.2) (2.8) (2.4) (3.1) (3.9) (2.6) (3.1) (3.4) (4.7)
(.9) (.9) (1.7) (1.7) (2.5) (2.8) (3.3) (3.6) (3.1) (3.7) (3.3) (4.8)
Data are expressed as number (percent) of participants in each group and were compared by the chi-square test or Fisher exact test. a Number of participants in intervention group are as follows: T1, 1,176; T2, 1,153; T3, 987; T4, 845. b Number of participants in control group are as follows: T1, 915; T2, 904; T3, 838; T4, 830. c There were 22 missing data points for this variable at T2, all in the control group. d There was one missing data point for this variable at T1 in the intervention group and 85 missing data points at T3 (77 in the intervention group and 8 in the control group). e There was one missing data point for this variable at T1 in the intervention group. *p < .05 indicates a significant difference between the two groups.
320
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Figure 2. Changes in the theory of planned behavior variables and life skills for participants in the intervention and control groups include the results of the linear mixed model. (A) ATT ¼ attitudes. (B) SN ¼ subjective norms. (C) PBC ¼ perceived behavioral control. (D) LS ¼ life skills. (E) BI ¼ behavioral intention not to use illicit drugs. T1 ¼ baseline; T2 ¼ after the main intervention program; T3 ¼ after the first booster session; T4 ¼ after the second booster session. *Significant difference (p < .05) between the two groups. Error bar indicates standard deviation.
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
intervention and two booster sessions, was implemented at 12 middle schools among students aged 13e15 years, and showed evidence of effectiveness as indicated by the analysis of findings from students’ self-reported questionnaires assessing illicit drug use and changes in TPB indicators and life skills from the beginning of the program until 12 months after main intervention completion. In our previously published pilot study of the TPB/life skills school-based intervention program [13], we found that participants who received the intervention (in 10 45-minute sessions) demonstrated superior performance in attitudes, subjective norms, perceived behavioral control, and life skills compared with participants in the control/nonintervention group. Furthermore, participants in the intervention group were also found to have significantly greater behavioral intention not to use illicit drugs than participants in the control group [13]. Although the present study was designed on the basis of the pilot study, some aspects of the present implementation were unique. First, the illicit drug prevention program was modified by adding two booster sessions at 6 and 12 months after the second assessment. Second, the two booster sessions allowed us to collect follow-up data at 6 months (T3) and 12 months (T4), partially representing the long-term effect of the intervention program. From the beginning of the program to the second booster session, the students were progressing from the eighth grade to the ninth grade. Third, the booster session indeed provided an opportunity to reinforce the topic(s), especially the topics that students had not mastered in the previous assessment. White and Pitts (1997) [23] suggested that booster sessions provide the opportunity to reinforce and build on messages over a number of years suited to the age and development of the students. It is notable that the prevalence of illicit drug use at the end of program was increased in the control group (from .9% at T1 to 1.7% at T4) but remained unchanged in the intervention group (from .3% at T1 to .2% at T4). This suggests that our illicit drug use prevention program has the potential to effectively prevent (prophylactic effect) high school students from using illicit drugs in adulthood. Consistent with this finding, participants in the intervention group also had significantly higher scores for TPB and life skill measures at the end of the booster sessions. These findings indicate that the incorporation of TPB and life skills into the intervention program helped promote participants’ attitudes and subjective norms, and may potentially facilitate reducing illicit drug use. Our findings are in keeping with those from previous TPB studies that have demonstrated associations between attitudes and intentions and consequent illicit drug (marijuana) use [24] and between attitudes and subjective norms and the intention to use illicit drugs [25]. The effectiveness of skills-based drug use prevention programs has been demonstrated by a systematic review. A systematic review of 29 randomized control trials on drug abuse prevention programs, of which the majority were conducted in the United States among the sixth and seventh graders just slightly younger than our target groups, found that skills-based interventions significantly reduced marijuana use and hard drug use, and improved decision-making skills, self-esteem, and peer pressure resistance compared with conventional curricula [16]. A key part of our intervention program was the integration of two booster sessions that were implemented at 6 and 12 months after completing the initial intervention. It is interesting that, for all domains tested in the questionnaire, the scores between
321
groups only showed significant differences at the end of each booster session but not at the end of the 10-session main program (the second assessment; T2). This observation may result from two possible reasons: (1) the second assessment (T2) was not performed at the end of the 10-session main program but at the end of the 8- to 10-week summer vacation. Students were given homework exercises to complete during their summer vacation, but they apparently did not take the homework seriously. Therefore, the intervention protocol may require revision to ask the students to complete the homework exercise immediately after the end of the 10-session program and (2) the lack of between-group differences immediately after the main intervention may simply be a time-related effect. For example, a previous study examining the effects of an online intervention program (without booster sessions) on subsequent drug use among girls in grades 7e9, found a significant difference in the rate of drug use between the intervention and control groups at 6 months after program completion but not immediately after the program [26]. The present study attempted to reduce possible biases and other confounders (e.g., cultural norms, expectations) by using an anonymous questionnaire. Students were informed before filling out the questionnaires that the outcome of the study would be for research use only, and only the research team in this program would be able to access student data. In addition, not all schools and students enrolled completed the study. As a consequence, the withdrawal of schools and students may have biased our findings to some extent. To rule out differential attrition as a threat to conclusion validity, we have conducted sensitivity analysis and attriter-remainder comparison analysis to confirm that differential attrition between the intervention and control groups in our study did not bias the study outcomes. Results of sensitivity analysis and comparison of attriter-remainder differences at different time points indicated that no significant differences existed between attriters and remainders in main outcome variables, confirming that the potential for bias due to differential attrition was minimal. Our study has a number of limitations that warrant mention. First, we did not compare the effectiveness of our intervention with other evidence-based drug prevention programs for middle school students such as Project ALERT in the United States [9] or the LifeSkills Training Program outlined by Botvin and Kantor [15]. Given that differences exist in cultural norms and practices, the program implemented in the present study may not be readily applicable to other populations and vice versa. Second, we cannot comment on the longer term effectiveness (i.e., a number of years later when the students become adults) of our intervention program. Additional follow-up is needed to determine how the described intervention program affects illicit drug use beyond 12 months after completion of the main intervention. In conclusion, after participants completed the main intervention and two booster sessions of the theoretically based drug use prevention program, the scores for attitudes, subjective norms, perceived behavioral control, life skills, and behavioral intention were significantly higher among students receiving the intervention than among the control groups receiving conventional antidrug use didactic teaching. These results suggest that our drug use prevention program that integrates life skills into the TPB may be effective for reducing illicit drug use and improving planned behavior-related constructs in adolescents. Pending further research and confirmation of the findings reported herein, practical implementation of this program may be warranted.
322
J.-L. Guo et al. / Journal of Adolescent Health 56 (2015) 314e322
Acknowledgments The authors thank all students and teachers who participated in the study. Funding Sources This work was supported by grants from the Taiwan Food and Drug Administration, Department of Health (DOH 99-FDA61209; DOH 100-FDA-61202). This article was subsidized by the National Taiwan Normal University, Taiwan, ROC. Supplementary Data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.jadohealth.2014.12.003. References [1] Degenhardt L, Hall W. Extent of illicit drug use and dependence, and their contribution to the global burden of disease. Lancet 2012;379:55e70. [2] Swendsen J, Burstein M, Case B, et al. Use and abuse of alcohol and illicit drugs in U.S. adolescents: Results of the National Comorbidity SurveyAdolescent Supplement. Arch Gen Psychiatry 2012;69:390e8. [3] Vega WA, Aguilar-Gaxiola S, Andrade L, et al. Prevalence and age of onset for drug use in seven international sites: Results from the international consortium of psychiatric epidemiology. Drug Alcohol Depend 2002;68:285e97. [4] Chen CY, Storr CL, Anthony JC. Early-onset drug use and risk for drug dependence problems. Addict Behav 2009;34:319e22. [5] Rubino T, Zamberletti E, Parolaro D. Adolescent exposure to cannabis as a risk factor for psychiatric disorders. J Psychopharmacol 2012;26:177e88. [6] Botvin GJ, Griffin KW. School-based programmes to prevent alcohol, tobacco and other drug use. Int Rev Psychiatry 2007;19:607e15. [7] Guo JL, Huang CM, Lee TC. National Web-based survey of illegal drug use among adolescents. Taipei, Taiwan: Ministry of Education; 2008. [8] Chen KT, Chen CJ, Fagot-Campagna A, et al. Tobacco, betel quid, alcohol, and illicit drug use among 13- to 35-year-olds in I-Lan, rural Taiwan: Prevalence and risk factors. Am J Public Health 2001;91:1130e4. [9] McBride DC, Terry-McElrath YM, VanderWaal CJ, et al. U.S. Public Health Agency involvement in youth-focused illicit drug policy, planning, and prevention at the local level, 1999-2003. Am J Public Health 2008;98:270e7.
[10] Des Jarlais DC. Prospects for a public health perspective on psychoactive drug use. Am J Public Health 2000;90:335e7. [11] Ellickson PL, McCaffrey DF, Ghosh-Dastidar B, et al. New inroads in preventing adolescent drug use: Results from a large-scale trial of project ALERT in middle schools. Am J Public Health 2003;93: 1830e6. [12] Griffin KW, Botvin GJ. Evidence-based interventions for preventing substance use disorders in adolescents. Child Adolesc Psychiatr Clin N Am 2010;19:505e26. [13] Huang CM, Chien LY, Cheng CF, et al. Integrating life skills into a theorybased drug-use prevention program: Effectiveness among junior high students in Taiwan. J Sch Health 2012;82:328e35. [14] Buhler A, Schroder E, Silbereisen RK. The role of life skills promotion in substance abuse prevention: A mediation analysis. Health Educ Res 2008; 23:621e32. [15] Botvin GJ, Kantor LW. Preventing alcohol and tobacco use through life skills training. Alcohol Res Health 2000;24:250e7. [16] Faggiano F, Vigna-Taglianti FD, Versino E, et al. School-based prevention for illicit drugs use: A systematic review. Prev Med 2008;46:385e96. [17] Dane AV, Schneider BH. Program integrity in primary and early secondary prevention: Are implementation effects out of control? Clin Psychol Rev 1998;18:23e45. [18] Botvin GJ. Preventing drug abuse in schools: Social and competence enhancement approaches targeting individual-level etiologic factors. Addict Behav 2000;25:887e97. [19] McBride N. A systematic review of school drug education. Health Educ Res 2003;18:729e42. [20] Rubio DM, Berg-Weber M, Tebb SS, et al. Objectifying content validity: Conducting a content validity study in social work research. Soc Work Res 2003;27:94e104. [21] Francis JJ, Eccles MP, Johnston M, et al. Constructing questionnaires based on the theory of planned behaviour: A manual for health services researchers. University of Newcastle upon Tyne: Centre for Health Services Research, Newcastle upon Tyne; 2004. [22] Cohen, J. Statistical power analysis for the behavioral sciences (2nd ed.). Hillside, New Jersey: Lawrence Erlbaum Associates;1988. [23] White D, Pitts M. Educating young people about drugs: A systematic review. Addiction 1998;93:1475e87. [24] Hohman ZP, Crano WD, Siegel JT, et al. Attitudes ambivalence, friend norms, and adolescent drug use. Prev Sci 2014;15:65e74. [25] Bashirian S, Hidarnia A, Allahverdipour H, et al. Application of the theory of planned behavior to predict drug abuse related behaviors among adolescents. J Res Health Sci 2012;12:54e60. [26] Schwinn TM, Schinke SP, Di Noia J. Preventing drug abuse among adolescent girls: Outcome data from an Internet-based intervention. Prev Sci 2010;11:24e32.