Two-year Effects of a Classroom-based Smoking Prevention and Cessation Intervention Program

Abstract

Background

Project EX has been well-evaluated for prevention and cessation effects among adolescents in Spain. However, its long-term efficacy is unknown. The current study provides outcomes of a two-year follow-up evaluation of Project EX.

Methods

The intervention was tested using a clustered randomized controlled trial with 1,546 Spanish students, involving three program and three control schools. At two-year follow-up, 722 subjects also completed the questionnaires (266 in the control condition and 456 in the program condition).

Results

Compared to the control condition, the program condition revealed a greater reduction in nicotine dependence (p = .04), smoking intention (p = .02), and number of cigarettes at last month (p = .03). The CO monitor repeated assessments revealed a significant decrease of ppm levels in the program group (p < .001). Intent-to-treat quit rates were 14.28% and 0%, respectively, for the program and control conditions.

Conclusions

This study provides evidence about the long-term effectiveness of the Project EX classroom-based program for adolescent prevention and cessation in Spain.

Tobacco consumption in adolescents remains an important public health problem. The World Health Organization (WHO) [1] estimates that 22% of adolescents over 15 years old are regular smokers of tobacco. Although tobacco use by adolescents has declined substantially in the last 40 years, in the United States nearly one in 15 high school seniors was a daily smoker in 2014 [2]. The latest survey from the European School Survey Project on Alcohol and Other Drugs [3] found that 54% of young Europeans aged 15 years had smoked cigarettes at least once in their lifetime, and more than half of the lifetime smokers had smoked cigarettes in the last 30 days (approximately 30% of all youth); meanwhile, in the United States only approximately 7% of high school students indicated some form of tobacco use at least once in the previous month [4]. In Spain, tobacco is the second most consumed addictive substance, with 35.3% of adolescents who have ever smoked [5]. The differences between countries are evident; tobacco control is one of Spain's most pressing public health needs. However, research and implementation is scarce.

Although the are important reviews on the efficacy of smoking cessation and prevention measures in adolescents [6-9], relatively few studies of teen smoking cessation have been conducted worldwide compared to adult cessation programs and only 25% of these studies have been conducted outside the United States [10]. Also, while some tobacco use-related prevention programming has been evaluated in Spain [11], only Project EX has been well-evaluated on the potential for prevention and cessation effects among adolescents in Spain [12, 13], and provides motivation-coping skills-commitment material [14]. Its immediate and one-year follow-up effects are known, but its long-term efficacy is unknown.

Project EX is an empirically-validated teen tobacco use cessation program, and provides motivation-coping skills-commitment material [14]. In its clinic version, Project EX has shown positive 3-to-6-month outcome effects over three experimental trials and one multiple baseline single group trial [15-17]. While the school-based clinic version of Project EX has been found to be effective in the U.S., China, Russia, and Spain, international translation research on Project EX thus far has been limited to those smokers who attend the clinic. The classroom version of Project EX sought to increase program reach and is delivered in the classroom setting, where both tobacco users and nonusers are located. A prevention/cessation classroom-based version of EX was delivered in southern California and revealed effects across the full range of smoking, and most strongly on cessation [18].

This paper focuses on the two-year self-reported behavioral outcomes of the Project EX prevention/cessation program with Spanish adolescents. In the previous paper on the implementation and immediate outcomes of the project, the program was well- received and produced immediate effects on reducing smoking intentions and CO ppm levels [12]. At one-year follow-up, the program condition revealed a greater reduction in nicotine dependence and CO ppm levels, and lower consumption of cigarettes at last month [13]. We hypothesized that the classroom-based curriculum would show maintenance of effects among the full sample at two-year follow-up, which would bolster the practical importance of program effects in Spain, as variations in sustained use or cessation are observed at the follow-up [19].

Methods

School Selection and Experimental Design

Before proceeding with the recruitment of the sample, the study was approved by the IRB at Miguel Hernandez University, ensuring that the study does not involve risks to participants, involves the implementation of informed consent, and maximizes confidentiality of the participants. The education authorities were informed of the study goals, and authorization was requested. Once such authorization had been obtained, the researchers interviewed the head teachers and the school counselors, informing them verbally and in writing about the aim of the study, so as to obtain their permission and encourage their cooperation. Finally, parents were informed by letter and requested to provide written consent for their children to participate in the study. The written parental consent was provided for all minors participating.

We contacted a convenience sample of 45 schools from 17 towns in the Province of Alicante. After the first in-person meeting with the school boards to present the objectives of the intervention, a total of six high schools from three cities (Elche [n = 4], Crevillente [n =1] and San Vicente [n =1]), were recruited (recruitment rate = 13%). The reasons for non-recruitment were no response (72%), responded back to us after a first meeting with a statement of no interest (18%) and inability to be able to implement the study during the school day (10%). The schools recruited were randomly assigned to one of two experimental conditions: treatment or standard care (control); that is, there were three schools per condition (schools were carefully matched into pairs prior to assignment on school size, and the randomization was at the level of the school) [20].

Project EX Curriculum

The original version of the EX prevention/cessation curriculum was translated into Spanish by two translators. Two bilingual researchers working at the University Miguel Hernandez (UMH) checked the translation by reading both the English and Spanish versions. The feasibility of the Spanish version was assessed in a focus group composed of ten students at a high school, before program implementation, which verified that the curriculum was clearly understood and culturally appropriate.

The Project EX classroom curriculum is closely adapted from the clinic program [21, 22]. The learning activities included strategies to not start or quit smoking and learning skills for cessation maintenance, with a motivation-enhancement, interactive protocol. The program comprised several activities including: 1) role playing (“talk shows”), where student volunteers fill the role of each guest and the audience (other students in class) direct questions to the guests; 2) alternative medicine strategies including healthy breathing, characterized by slow and deep breaths, and yoga and meditation activities, which involves focusing attention on the present moment; and 3) a game, “is smoking on the menu”, which divides the participants into teams to answer questions about tobacco. The curriculum involves eight sessions delivered in classrooms (50 minutes each session) over a six-week period. The first four sessions are held in a two-week period. The second four sessions are held once per week during the following month. Session 1: Orientation imparts the ground rules for the class and discusses reasons for using, not using, quitting tobacco, or remaining tobacco free. Session 2: Tobacco affects your life, discusses how tobacco use can cause, rather than relieve stress. Session 3: Health dangers of tobacco use, discusses the harmful substances in tobacco and how it can injure one's body. Session 4: Quitting step 1-Making a commitment about not using tobacco, discusses addiction to tobacco. Methods of quitting and physical and psychological aspects of withdrawal are discussed. Session 5: Quitting step 2-Managing withdrawal symptoms, discusses more about nicotine, addiction, and strategies of avoiding addiction or managing withdrawal symptoms. Psychological coping includes self-forgiveness and avoiding false expectations regarding how not using tobacco or quitting will and will not affect one's life. Session 6: Taking care of a healthy body, involves learning lifestyle balance strategies, including weight control and practicing a “yoga activity”. Session 7: Taking care of your piece of mind, involves learning more coping strategies, including assertiveness training and anger management. Participants also learn the “letting feelings pass” meditation activity. Finally, Session 8: Not smoking again: commitment and avoiding relapse, involves learning means to avoid using tobacco again, or staying tobacco free, and mentions how topics covered in the tobacco education program could be applicable to other substances. More detailed description of the sessions in the previous paper on the implementation and immediate outcomes of the project [12].

Cultural Adaptations

In addition to language adaptations, five changes were made in the curriculum to adapt it to Spanish culture. First, the original curriculum targets “tobacco” use including smokeless tobacco, pipes, cigars, and chewing tobacco. Spanish adolescents very rarely use other forms of tobacco other than cigarettes [22, 5], so that information was removed and the program focused on prevention or cessation of cigarette smoking only. Second, when Project EX was implemented as a research project in the United States, extrinsic motivators were provided as incentives. Participants were told that they would obtain credits for participating in the program. In Spain, as in other international settings within which Project EX is being translated [15,16], there were no incentives for attendance at the program sessions. Third, all names of characters in the talk [22] were changed from American to Spanish names. Fourth, the monetary amounts were changed from dollars to euros. For instance, in Session 1, “Orientation”, the original curriculum provides as a suggestion for the question “What are some reasons for quitting?” that one may “Save money. Over US $2,008 per year is spent on smoking one pack per day”. The amount was changed to 1,249 euros, which indicates what the cost is in Spain. Finally, in Session 3, “Health Dangers of Tobacco Use”, there is a question list for a game in the original curriculum, pertaining to second-hand smoke and policies. Questions and forced-choice responses were changed to reflect policies in Spain. For example, one question asks “In which of the following places is it legal to smoke in the United States? (‘Airline flight,’ ‘Interstate bus,’ ‘New York City taxi cab,’ and ‘None of the above’)”. This item was replaced with “In which of the following places is it legal to smoke in Spain? (‘Airport,’ ‘Bus station,’ ‘Taxis,’ and ‘None of the above’ [correct answer])”.

Training in Project EX

Between October and November of 2012, psychology graduate students at Miguel Hernandez University were given the opportunity to become a program facilitator. All volunteers received an introductory lecture about Project EX, which included a brief summary of the Project EX curriculum, its history, advantages of using the program, and the role of program facilitators. Seventeen Spanish graduate students (16 female), an average of 23 years of age (SD = 0.8), interested in implementing the program, spent eight hours studying and practicing the eight sessions of Project EX, including learning details of program delivery and how to deliver the material with fidelity. The training was provided by a researcher who had previously been trained by the program developer. All persons trained delivered the program, an average of two classrooms each.

Participants

An average of 12 classes was selected per school, with a range of 8 to 19 classes, across the six high schools. A total of 1,546 students were selected for participation in the study (716 in the control condition and 830 in the program condition). Among the 1,546 subjects who participated in the pretest survey, 965 (393 in the control condition and 572 in the program condition) also completed one-year follow-up questionnaires (62.4% retention rate). At two-year follow-up, 722 subjects (266 in the control condition and 456 in the program condition) also completed the questionnaires (46.7% retention rate). All adolescents who conducted assessments of baseline and 24 months follow-up were included in the analysis. Among the 722 subjects who participated in the two-year follow-up questionnaires, 440 also completed the CO readings.

Participants varied from 14 to 20 years of age (M = 15.17; SD = 1.20). The sample was 49.6% male (n = 358); 93.1% Spanish (n = 672), and 6.9% (n = 49) other nationality (e.g., South America, Morocco, Eastern Europe). Further, 81.2% (n = 586) of the students lived with both parents, and 36.6% (n = 264) of youths' fathers and 37.5% (n = 270) of youths' mothers completed high school. Approximately 30% (n = 216) of the participants had smoked a cigarette sometime in their lives and 15.2% (n = 109) in the last 30 days, and 5.3% (n = 38) on the day of the pretest assessment.

Data Collection and Measures

Pretest and two-year follow-up measures were collected from students using a self-report, closed-ended and fill-in-the-blank response questionnaire, which were administered over one class period. At the two-year follow-up, students who failed to return the survey were contacted by telephone for survey administration. Demographic items included age (in years), gender, nationality (born in Spain, or immigrated to Spain from another country), current living situation (with parents, live alone, other situation), and parents' education (mean response across father's [or stepfather's] and mother's [or stepmother's] educational levels) based on categories derived from Hollingshead and Redlich [23].

Smoking behavior was assessed with the fill-in-the blank item asking “How many cigarettes have you smoked in the last month (30 days)?”. The level of nicotine dependence was assessed with the 8-item modified Fagerstrom Tolerance Questionnaire (mFTQ) [15, 24]. The total score is the sum of scores on the items (range from 0 to 25). To assess smoking intention [25], students were assessed with the question “How likely is it that you will smoke cigarettes in the next 12 months?”, with response categories being “1: Definitely not”, “2: Probably not”, “3: A little likely”, “4: Somewhat likely”, and “5: Very likely”. Furthermore, to validate assessment-day smoking responses, assessment of expired CO was completed at pretest and two-year follow-up by use of a breath CO monitor (Micro+ Smokerlyzer; Belfont Technical Instruments, Kent, UK; http://www.bedfont.com/ch/smokerlyzer/micro, accessed 4-19-2014). Participants who responded to our two-year follow-up telephone survey (60.94% of the total sample at two-year follow-up) did not complete the CO readings.

Data Analysis

A total of 1,546 students had been selected for participation in the study (716 in the control condition and 830 in the program condition). Among the 1,546 subjects who participated in the pretest survey, 965 (393 in the control condition and 572 in the program condition) also completed one-year follow-up questionnaires (62.4% retention rate). All adolescents who conducted assessments at baseline and 12 months follow-up were included in the one-year analysis [9]. At 2-year follow-up, only 722 participants were surveyed at pretest and that follow-up (46.7% of baseline subjects). To assess the potential sampling bias due to attrition at the two-year follow-up across conditions (external invalidity), we calculated a new variable called “attrition group” in which we included 722 participants who were surveyed at both time points and compared them with the 824 participants who were only surveyed at pretest. We used logistic regression analysis with “attrition group” as the dependent variable. Predictors included age, gender, nationality (born in Spain, or not), current living situation, parents' education, last month smoking, level of nicotine dependence, and smoking intention. To assess the potential sampling bias due to attrition at the two-year follow-up as a function of condition (internal invalidity), comparisons were made between the sample that was lost at two-year follow-up and the rest of the sample that remained in the study at two-year follow-up, as a function of condition with these same predictors. The comparisons utilized were chi-square or t-test models to indicate statistically significant differences (two-tailed p value at the .05 level). Again, all relevant demographic and outcome variables were examined.

Data analysis for four outcomes variables, which compared the two conditions, was completed with multi-level analysis [26] by using IBM SPSS Statistics 21 [27]. The outcome variables evaluated in this analysis were smoking intention, last month smoking behavior, nicotine dependence (using mFTQ score), and CO readings. Condition was considered a fixed effect variable; fixed at desired experimental levels (school, considered as a random effect variable). The variables adjusted for in the analyses included baseline measurements for each respective outcome variable, age, gender, status of living with both parents, and attrition propensity. The propensity for attrition score was calculated from a model predicting the actual attrition status with pretest measures [28]. The pretest measures in the attrition propensity prediction models included age, gender, ethnicity, whether or not participant lived with both parents, program condition, and last month cigarette use. Two-tailed significance tests were employed for the significance level calculation in all analyses (see Table 1). In an additional analysis, intent-to-treat (ITT) quit rates were calculated for those who said that they had smoked in the last 30-days on assessment day at baseline.

Table 1.

Program Effects Adjusted for Imbalance in Attrition ^a.

Outcome variables	Program (n =456)			Control (n =266)			Net Effect (Change in Experiment - Change in Control)
							Net	95%CI		p (df=5)
	Average	SE	p	Average	SE	p		Lower	Upper	p Value
Smoking intentionb	1.43	0.08	<.001**	1.99	0.27	.36	-0.56	-1.15	-0.24	.02*
Smoking last month c	0.56	0.18	.27	1.54	0.66	.52	-0.98	-0.36	0.42	.03*
mFTQ Nicotine Dependenced,e	0.56	0.07	.45	0.29	0.24	<.001**	0.27	-0.28	1.32	.04*

Notes:

^aAdjusted for the specific outcome assessed at baseline, age, and propensity for attrition center is modeled as a random effect.

^b5-Point scale from Definitely not (1) to Very likely (5).

^cNumber of cigarettes smoked in the last 30 days.

^d5-Point scale from Yes, I already have (1) to No (5).

^eModified Fagerstrom Tolerance Questionnaire score.

^*p < .05, two-tailed

^**p < .001, two-tailed

Results

Assessment of Attrition Bias at Two-year Follow-up

First, regarding the analysis of external invalidity, differences were only found on age (mean baseline age of group lost to follow-up = 15.28 years; mean baseline age of group surveyed at both time points = 15.17; p < .01). The attrition analysis comparing conditions (internal invalidity) revealed statistically significant differences in baseline age, smoking intention, and nicotine dependence between those who were retained at two-year follow-up and those who were not. The participants who were followed-up at two-year follow-up in the program condition at baseline were more likely to intend to not smoke in the future than in the control condition (73.2% versus 18.3%; t_(1,429) = 6.72; p < .05), and had less nicotine dependence than in the control condition (63.4% versus 31.1%; t_(1,932) = 13.56; p = .001). Differences in these variables have been controlled for in the analysis through use of the propensity score.

Effects on Smoking Intention and Tobacco Use

As shown in Table 1, the program was found to be statistically significant in reducing smoking intention, nicotine dependence, and number of cigarettes at last month. Subjects in the control condition reported a higher smoking intention at the two-year follow-up (p = .02), a higher level of nicotine dependence (p = .04), and increased consumption of cigarettes (p = .03). Further, at two-year follow-up the CO monitor repeated assessments revealed a significant decrease of ppm levels in the program group (-0.76, p < .001), while the control group showed a significant increase of ppm levels (0.37, p < .001).

The intent-to treat quit rate comparison was statistically significant (p < .05), among those reporting having smoked at baseline. Intent-to-treat quit rates were 14.28% for program condition (16/456) and 0% for control condition (0/266). Furthermore, the baseline non-smokers did not initiate smoking at any wave of assessment (posttest, one-year, two-years) in either condition of the study.

Discussion

The high rates of teenagers smoking in Spain indicate the need to develop effective long-term programs for tobacco prevention and cessation. The Project EX classroom curriculum was designed to produce smoking prevention and cessation effects in the same program. This study tested the two-year effects of the prevention/cessation program among Spanish adolescents.

The present study is the first to evaluate the two years effects of Project EX anywhere, showing significant differences between program and control condition on smoking intention, nicotine dependence and last 30-day cigarette use, in addition to a significant decrease of ppm levels in the program group along with an increase in ppm levels in the control group. These results confirm the hypothesis from the previous studies [12, 13], in which we expected to find longer-term effects of the program.

Considering previous studies of Project EX, the results presented in this work are promising. About Project EX tobacco use cessation, four-month follow-up data in China indicated a 10.5% 30-day quit rate and a 14.3% 7-day quit rate, compared to a 3% naturally occurring quit rate in a multiple baseline design [29]. The same program in Russia [15] significantly reduced future smoking intention and intention to quit, and increased motivation to quit smoking at immediate posttest. The program resulted in a 7% quit rate at the 6-month follow-up, compared to a 0% control group quit rate in a randomized controlled trial. In Spain, immediate outcomes of Project EX showed a reduction in future nicotine dependence scores and an increase of intention to quit smoking. At six-month follow-up Project EX had a significant influence on future smoking expectation and overall level of 30-day smoking, and approximately 14% quit in the program group versus 0% in the control group in a clustered randomized controlled trial [12].

In the classroom context, Project EX in southern California produced a greater reduction in weekly smoking and intention for smoking in the next 12 months [20]. At 6-and-12 month follow-ups students in the treatment condition experienced a greater reduction in weekly smoking and monthly smoking [16]. In addition, an incremental 8% higher cessation rate was observed at the program schools [13]. In Spain, immediate outcomes of Project EX [12] showed a greater reduction in smoking intention and CO ppm levels, and one-year follow-up outcomes revealed a greater reduction in nicotine dependence and CO ppm levels, and lower consumption of cigarettes at last month [13].

The long-term results of Project EX in Spain are relevant to clinical researchers and promoters of health education in schools, with important implication for the design of efficacious interventions to prevent and reduce tobacco use among adolescents over a relatively long time span.

Limitations

Project EX as a prevention-cessation program can effect long-term changes. Still, there are several notable limitations. First, recruitment rate of schools was 13%. Readiness to quit smoking is relatively low among Spanish adolescents and there is a low motivation among staff to add extra-curricular activities into the school the schedule. We were not able to offer the schools any incentives for participation (e.g., monetary donation to the school). Second, the dropout rate at two-year follow-up questionnaires was 53.3%. However, we did make at least three attempts per school to collect follow-up data, suggesting to us that higher retention rates would be difficult to obtain. Student absentees accounted for most of the non-completion though subject withdrawal from the study accounted for 47% of the non-completion of questionnaires at follow-up. Students who failed to return the surveys were contacted by telephone for survey administration; only 31% of persons called were reached and surveyed. Third, subjects who completed the survey by telephone did not complete the CO readings. However, concordance of results between those who completed the survey and CO measure suggests that youth reported honestly. Finally, U.S. and international survey data reveal that youth are aware of e-cigarettes and use of these products in this population is rapidly increasing [30], currently unregulated [31, 32]. In this study conventional tobacco consumption is evaluated, only. Future studies on the consumption or cessation of tobacco should consider using this popular type of electronic nicotine delivery system.

Recruitment and retention were major issues in the study. There appears to exist a low level of readiness at the school-level to be involved in teen prevention/cessation programming in Spain. We speculate that schools simply were too overloaded with structured programming to take on a research study. It is also possible that there is a reluctance to be involved in research, or that tobacco use is perceived as low priority at the schools in the Province. Possibly, incentives are needed to recruit schools and retain participants (particularly older participants who may be relatively likely to have competing commitments), or perhaps a longer recruitment process is needed to mobilize schools against tobacco use. Future research needs to examine these possibilities in more detail. We need to learn how to better engage schools in in youth tobacco use cessation research and practice. Despite the limitations of this study, the use of a randomized design (albeit only three units per condition) with a good baseline comparability, and the large sample size, the moderate likeability of the program [12], along with changes observed particularly on level of 30-day smoking and CO levels, suggests that there may be promise of Project EX as a cessation/prevention program for Spanish adolescents. Furthermore, there are no well- evaluated cessation intervention programs for adolescent tobacco in Spain, so the results of this study provide evidence about the long-term effectiveness, having a single program covering the entire Spanish adolescent population (both smokers and non-smokers), potentially more effective for society.