Study Design. Following screening and giving informed assent/consent, eligible participants will be randomly assigned to one of two study conditions: (1) MI: Two hour session of therapist administered MI, and (2) PE: Two hour session of therapist assisted viewing and discussion of educational DVDs about adolescent alcohol use, drug use, and driving under the influence provided by Human Relations Media, Mount Kisco, NY (hrmvideo.com).
Interventions will be curriculum based in that each session will be introduced by the leader with a brief description of problems associated with underage drinking, underage binge drinking, underage tobacco and other drug use, and driving while intoxicated. Participants will be randomly assigned to the MI or PE condition, but will be able to choose an individual or group intervention format. In a preliminary study, the investigators have found that many youth prefer a group format. Both individual and group formats have been successfully used previously in this population.
Assessments will be carried out pre-intervention and once within 18 months post-intervention. Pre-intervention assessments will be carried out in person and post-intervention assessments will be done in person, by telephone, or be web-based using questionnaires developed at Brown University and previously used successfully in youth in this population. Participating youth will be paid $75 for the initial two hour intervention and $75 for completing the follow-up. The payments will be in the form of Target or other department store gift cards. All questionnaires will be available on scannable paper forms or as a web-based application to permit ease and accuracy of data transmittal to a database for evaluation.
Measures: In addition to collecting demographic data, the following measures will be used pre-intervention (MI or PE) and at follow-up: Adolescent Drinking Questionnaire (ADQ), Drug Use Questionnaire, Drinking and Drug Use Driving Scale, Tobacco Use Scale, Injury behavior checklist, Student Self Check, Relationship with Parent/Guardian - Most Important, Teen Monitoring (Chassin), Peer Substance Use, Peer Tolerance of Substance Use, Stage of Change Assessment, MI Adherence Checklist (for the MI sessions only).
Examples of potential factors include information transfer, change in beliefs, internalizing new norms, resolving ambivalence, establishing increased autonomy, empowering for change, peer modeling, and peer support. The investigators will assess 1) these change factors, 2) the participant's movement along the stages of behavior change to a new level of drinking/drug use, and 3) self-reported changes in drinking, drug use, and intoxicated driving. With these data the investigators will be able to determine not only the overall efficacy of the MI vs. PE intervention in decreasing alcohol and drug use and intoxicated driving, but also the type of behavior change (e.g., moving from pre-contemplation to action) and the variables associated with that change (e.g., internalizing new norms, increased autonomy, peer support).
The MI intervention will incorporate open-ended questions, personalized feedback, and discussion about participants' alcohol use and drug, associated risk behaviors (e.g., drinking and driving), and the consequences of these behaviors. Individual MI procedures will incorporate the core principles of MI described by Miller and Rollnick, including expressing empathy, developing discrepancy, rolling with resistance, and supporting self-efficacy. Therapist interventions will be tailored to the participants' readiness to change/current stage of change (pre-contemplation, contemplation, preparation, action, maintenance, and relapse).
In the PE intervention, participants will view all four DVDs (all DVDs available from Human Relations Media):
Session 1: Underage Drinking: Know the Facts, Know the Risks describes negative consequences of youth alcohol abuse, including deaths resulting from drunk driving, and alcohol poisoning.
Session 2: Too Much: Extreme Danger of Binge-Drinking presents information specifically about the physiology of how binge drinking can cause death, and an interview with a parent of a student who died from alcohol poisoning while partying with his friends.
Session 3: Everything You Need to Know About Substance Abuse in 22 Minutes outlines alcohol, tobacco, and other drugs' effects.
Session 4: Dying High 2: Real Stories of Drugged Driving shows consequences of driving under the influence.
Training in MI. Under the supervision of Dr. Gilder, at least six training workshops will be held with Tribal clinic study staff to deliver culturally appropriate MI to adolescent Native Americans. Workshop leaders will be Motivational Interviewing Network of Trainers (MINT) certified PhD-level trainers with expertise in working with American Indian/Alaska Native populations.
Assessment of MI fidelity. Since tribal leaders and parents, citing confidentiality and cultural concerns, stated that audio- and video-taping of MI sessions would not be acceptable, the investigators will assess the MI sessions with the Adherence Checklist - Individual and Group MI Prevention Session. The checklist contains 14 items which assess the session's process and content and the therapist's adhering to the principles of MI. Items include asking if the therapist discussed the pros and cons of changing behavior, and introduced a change plan worksheet to the participant. Therapist fidelity to MI will be assessed both during training and on an ongoing basis every 6 months during the study using the MI Adherence checklist. If the average total checklist score of adherence falls below 80%, a quality improvement program will be instituted to identify the problem(s) and correct it with additional training.
Follow-Up. The investigators believe, given the success of the preliminary pilot study in recruiting reservation youth to the initial assessment and intervention, that successful follow-up of participants represents the main potential problem for the study. These obstacles were overcome, at least in part, by 1) instituting a web-based assessment that adolescents can take any time, including free time at school; 2) keeping email and regular mail addresses of participants current; 3) increasing payment for follow-up; 4) instituting follow-up in conjunction with a program the adolescent is participating in, i.e. after school youth program, and appointments at the clinic for medical and dental services. The Tribal IRB has given permission for the web-based assessment.
This intervention study aims to recruit 150 participants to each intervention condition for a total enrollment of 300 youth participants ages 13-20 years. Evaluation will be designed to test that MI is more effective than the PE in reducing alcohol and drug use and intoxicated driving at follow-up. Evaluations to test that MI is more effective than the PE in reducing alcohol and drug use will be undertaken as follows: The primary outcome variables will be: in the previous 6 months (1) quantity x frequency of drinking, (2) frequency of binge drinking and being drunk; (3) number of days used cigarettes, other forms of tobacco, e-cigarettes and nicotine inhalant devices, cannabis, "Spice," stimulants, opiates, cough syrup, inhalants, hallucinogens, and benzodiazepines; (4) drinking or using drugs and driving and/or being a passenger in a car with a driver who has been using alcohol or drugs; (5) a negative consequences of drinking or using drugs composite score.
Descriptive statistics will be generated for each variable. Statistical analyses assessing MI vs. PE differences will be assessed as follows. For dichotomous variables, repeated-measures chi-square analyses will be used. For continuous variables, repeated measures analysis of variance (ANOVA) or analysis of covariance (ANCOVA) will be used. The Mann-Whitney Test will be used for continuous variables that are not normally distributed. Repeated measures of ordinal data for two time points will be assessed using the Wilcoxin Signed Ranks Test. With 2 groups, 150 participants in each group, an alpha = 0.01 to avoid a Type 1 error, a medium effect size for the group of dependent variables analyzed in chi-square, ANOVA, and logistic regression analyses, and a Bonferroni correction for 5 multiple comparisons in each set of analyses, the power at the projected n for chi-square, ANOVA, and logistic regression will be 0.96, 0.96, and \>0.80, respectively. Power estimates were made using G\*Power 3.61. If sample size permits, a larger model examining the association of selected outcome variables (e.g., frequency of bingeing and drinking/driving in the past 6 months) with independent variables (e.g., age, gender, stage of change, MI vs. PE intervention) in multiple and logistic regression models will also be evaluated.
