Health Education Research, Vol. 14, No. 6, 791-802,
December 1999
© 1999 Oxford University Press
Effectiveness of a social influence approach and boosters to smoking prevention
Department of Health Education, and
1 Department of Methodology and Statistics, University of Maastricht, PO Box 616, 6200 MD Maastricht, The Netherlands and
2 Center for Health Promotion Research and Development, University of Texas, Houston, PO Box 20188, Houston, TX 77223, USA
Correspondence to: I. Mesters
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
This paper presents the short-term and long-term results of a randomized smoking prevention trial. The purpose was to evaluate two smoking prevention programs, a social influence (SI) program and a SI program with an additional decision-making component (SIDM). Moreover, the contribution of boosters was assessed as well. Fifty-two schools were randomly assigned to the SI program, the SIDM program or a control group. Half of the treatment schools were randomly assigned to the booster condition; the other half did not receive boosters. Both programs consisted of five lessons, each lasting 45 min, and were given in weekly sessions in grades 8 and 9 of high schools in the Netherlands. The most successful program was the SI program with boosters which resulted in a significantly lower increase in smoking rates (5.6 and 9.7%, respectively) compared to the control group (12.6 and 14.9%, respectively) at both 12 and 18 months follow-up. The results suggest that boosters can be an effective tool for maintaining or increasing the effectiveness of smoking prevention programs. It is recommended that the SI program with the booster be implemented at the national level, since this intervention showed the greatest behavioral effects.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Among the antismoking education programs available today, the comprehensive social influence (SI) programs appear to be most successful in reducing the onset of smoking (Sussman et al., 1995). SI programs are based on the assumption that inoculation to resist social pressures that serve as precipitants of use would help to prevent use (McGuire, 1964). Applied to smoking, this model posits that resistance to persuasion will be greater if one has developed arguments with which to counteract pressures to smoke (Evans, 1976). Two prevention studies by the Houston group (Evans et al., 1978
Evans et al., 1979) were based on this theory, but also included insights derived from attitude change (persuasive communication) theory (McGuire, 1969) and social learning theory (Bandura, 1977). Other researchers have elaborated on the Houston studies, adding elements to SI programs, such as skills training techniques, commitment or peer teaching. Moreover, later studies used superior methodology to many of the earlier studies (Flay, 1985; Sussman et al., 1995).
An initial study of smoking prevention among students in the Netherlands found significant preventive effects of a SI peer-led program for pupils following vocational education (De Vries et al., 1994). The same study suggested that students receiving education at the higher educational level had a need to discuss the decision-making process more elaborately and wanted material at a higher reading level than the students in the vocational track (De Vries et al., 1994). This finding indicated that different smoking prevention programs were required for different educational tracks (De Vries, 1989b; Chatrou, 1992). Hence, two adjusted SI smoking prevention programs were developed for students following the higher educational tracks (De Vries, 1989a). In one program, called the decision-making program (SIDM), a distinct decision-making component was added to the SI approach. This decision-making component was based on the decision-making stages described by Janis and Mann (Janis and Mann, 1978). The decision-making component was included to investigate the additional effect of adding the decision-making component to the SI approach to smoking prevention.
The theoretical model underlying the present SI and SIDM programs is based on social inoculation theory (McGuire, 1969; Evans, 1976), the model of planned behavior (Ajzen, 1991), containing attitudes and social norms, as well as Bandura's self-efficacy expectations (Bandura, 1986), as important factors to guide the development of the program's content and/or measurement instruments to establish the impact of the programs. Besides social norms as determinant of smoking, the model also included the influence of modeling and social pressure on smoking behavior. The combination of constructs mentioned (attitudes, social norms, social pressure, modeling and self-efficacy) is referred to in the Netherlands as the ASE model (Attitude, Social influence and Efficacy model) (De Vries et al., 1994). According to this model, attitudes, social influences and self-efficacy expectations predict the intention to perform the behavior which, in turn, predicts the behavior itself. External variables, such as demographic variables, influence behavior via the three determinants and the intention. Between intention and behavior there may be barriers or lack of skills which may hinder the realization of the intention.
At the start of this project, in 1989, boosters were recommended in the literature in order to improve program results (Murray et al., 1989). At that time, boosters had rarely been tested for their additional effect in an educational program in a randomized trial. Recent studies have shown that boosters might actually fulfill their promises (Botvin et al., 1990). Today, various kinds of booster have been introduced, such as telephone interventions (Elder et al., 1993), mail boosters (Elder et al., 1993) or additional class room lessons (Ellickson and Bell, 1990). For the Netherlands, a non-labor-intensive type of booster had to be developed to ensure that the intervention could be implemented nationwide by a national organization for smoking prevention. Additional classroom lessons were not an option, because of the tight curriculum (De Vries, 1989b; De Vries et al., 1992), while telephone calls were considered too labor-intensive for a national organization. In this study, therefore, the boosters consisted of three magazines on smoking prevention that could be handed out to students by teachers at school.
Epidemiological data for 1989 (at the start of the study) indicated that the higher educational tracks showed lower incidence figures for smoking. For instance, 15% of the 12-year-old vocational-school students reported having smoked in the past 4 weeks, compared to none of the students of similar age following the highest educational track (Dutch Smoking and Health Foundation, unpublished data). Therefore, the new SI programs for the higher educational tracks were developed for an older age group, to respond to the later onset of smoking, and were consequently implemented in a higher grade (starting in grade 8, 13–14 year olds, and following through grade 9, 14–15 year olds).
This paper describes both short-term and intermediate results of a large-scale randomized smoking prevention trial among students following education at the intermediate and highest level. The schools in the study were assigned to the following conditions: (1) a peer-led intervention (SI or SIDM) with a booster program, (2) a peer-led intervention (SI or SIDM) without a booster program and (3) a control group. It was expected that the SIDM program would be more effective than the SI program in preventing and delaying smoking for students following higher education. Further, it was hypothesized that the programs with boosters would be more effective than programs that did not included them. Finally, we expected that the programs would result in increased knowledge about the effects of smoking, less favorable attitudes towards smoking, higher self-efficacy levels for non-smoking, a (more) negative SI regarding smoking and a (more) negative intention to smoke.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample
In the Netherlands 62 district health centers exist of which 20 were approached for participation in the study. Fifteen of them agreed to participate in the research project. Health educators working in the participating district health centers approached every school in their region to participate in the project (Dijkstra et al., 1993
). They obtained formal consent from school boards, since this is the standard procedure in the Netherlands. No parental consent was needed (De Vries et al., 1992). Fifty-two schools from 15 district health centers were randomly assigned by the university research team to the SI program (51 classes), the SIDM program (64 classes) or the control group (67 classes). The treatment condition resembles 32 schools, the control condition contains 20 schools. Within the treatment condition, half of the schools were randomly assigned to the condition receiving three boosters, while the other half did not receive any boosters. The interventions were implemented in grades 8 and 9.
Design
In the Netherlands, school starts around September and ends in July. The pre-test (T1) was late October 1990, program implementation took place during November and December (grade 8). The first post-test (T2) was in April 1991. The first booster was given late September 1991 (grade 9). The second post-test (T3) was in October 1991, and the second and third boosters were given beginning January and late March 1992. The boosters were disseminated with a 3 months interval. In April 1992, the third post-test (T4) was conducted. Program effects on smoking at T2, T3 and T4 were described. All those respondents who participated at any post-test survey and who could be matched with pre-test data were included. This resulted in overlapping but not identical samples at T2 (N = 4060), T3 (N = 3653) and T4 (N = 3104).
Elements of the SI program
Procedure
Peer-led activities were conducted in small groups consisting of four or five students. Before the first lesson, groups were formed and peer-leaders were chosen by the students themselves. The peer-leader was a non-smoking student from the same class as the students (Evans et al., 1978; Perry et al., 1987; De Vries, 1989a) and served as a chairperson for the small activity group. Teachers coordinated the lessons, stimulated students and assisted peer-leaders (Arkin et al., 1981).
Manuals and video for peer-leaders and teachers
Teachers received 1 h of training from health educators, consisting of information about the structure and content of the program, and how to stimulate students. To be successful, peer-leaders had to receive adequate training as well (Glynn, 1989). This training was done by the teacher. For this purpose a training video was developed in which the task of peer-leaders was explained. In addition to this training on video, the program consisted of separate manuals for teachers and peer-leaders, informing them about activities and their tasks.
Student manuals and video
The SI program consisted of five lessons, each lasting 45 min, and were given in weekly sessions. Structure and content of the program were presented on video. The structure can be summarized as follows: (1) introduction of the theme on video by adolescent models (Perry et al., 1983), (2) peer-led activity as indicated in the student manual (Klepp et al., 1986), (3) feedback on the activity provided on video and continuation of the lesson on video, (4) peer-led activity guided by the student manual, (5) feedback on the activity on video, and (6) home work assignments as described in the student manuals. The content of the lessons was as follows. The first lesson focused on the reasons why people do or do not smoke and why people quit smoking, and the differences between direct and indirect pressure to smoke. The second lesson dealt with the short-term effects of smoking, the dangers of experimentation with smoking, passive smoking, addiction and quitting smoking. Students received a brochure about how to quit smoking, which they could hand out to a smoking person at home or in their environment (Mudde et al., 1994). The third lesson focused on resisting peer pressure and acquiring skills to resist pressure. The fourth lesson discussed how to react when bothered by smoke, indirect pressure to smoke from adults and advertisements, and measures from the government against smoking. The last lesson focused on alternatives to smoking, making the decision to smoke or not and a commitment to non-smoking behavior.
Boosters
Three magazines were developed, discussing information similar to that contained in the SI video. Teachers were asked to distribute these magazines among the students, who could read them at home or during a break at school. In the three magazines well-known national and international singers and sports personalities served as non-smoking models and gave their opinion on smoking. There were interviews with non-smoking actors from the video and with a Greenpeace employee. Information was given on the effects of smoking, passive smoking, helping other people to quit smoking and on reasons for not smoking. Furthermore, each magazine included a competition, a cartoon about smoking and letters to the editor from students.
Additional materials
At the end of each lesson, teachers handed out a written summary of the lesson, which could be added to the manuals. To increase commitment to non-smoking, students were asked to conclude a non-smoking contract (anonymous commitment), and to write their name on a non-smoking poster that could be clearly seen in the school and consequently by other students (public commitment). As a reward for their non-smoking, non-smokers received a non-smoking poster.
Decision-making component
The student manual for the first lesson of the SIDM program discussed five steps towards making a decision. These five steps were based on the five decision-making stages described by Janis and Mann (Janis and Mann, 1978), who based the five stages on studies of people who displayed vigilance in reaching a difficult personal decision that they subsequently carried out successfully, such as giving up smoking, losing weight on a low-calorie diet or undergoing a prescribed medical treatment. Janis and Mann's five stages are: (1) appraising the challenge, (2) surveying alternatives, (3) weighing alternatives, (4) deliberating about commitment and (5) adhering despite negative feedback. In the first stage, a person is exposed to information about a threat or opportunity that effectively challenges a current course of action. This challenging information can be an event that disturbs the person's equanimity because a particular threat can no longer be ignored. Once the decision maker gives a positive response to the challenging information, he proceeds to search for alternatives. Stage 2 is largely devoted to discovering and selecting viable alternatives. In stage 3 the decision maker proceeds to a thorough search and evaluation, focusing on the pros and cons of each of the alternatives. The goal is to select the best available course of action by weighing the advantages and disadvantages of each alternative. After the decision maker has decided about the best alternative, he begins to deliberate about implementing it and conveying his intentions to others. In the last stage, the decision maker implements his choice. In the present smoking prevention program, students were asked to pass through the following process: (1) what is the situation in which you have to make a decision?, (2) what are the possible decisions? (3) what are the pros and cons of the possible decisions?, (4) make a decision based on the pros and cons, and (5) implement the decision.
Data collection procedure
Questionnaires for the effect and process evaluation among students were distributed by teachers. Respondents were assured of anonymity and confidentiality, and that no one but the researchers had access to the data. After respondents had finished their questionnaire, teachers put the papers in an envelope, sealed it in the classroom and sent it via the health educators to the researchers.
After the last lesson of the program, the teachers (N = 59) from 32 intervention schools received a questionnaire about the implementation and use of the program. They were asked to finish the questionnaire before the health educators started to conduct interviews with teachers to obtain more in-depth information about the use of the program [results of the interviews are described elsewhere (Dijkstra, 1995)].
Process questionnaire for teachers
Although differences in level of implementation of program elements between classrooms will be checked using multilevel analysis, it is interesting from a program developers point of view to know to what extent program elements have been used. The level of implementation of the smoking prevention programs was assessed by asking teachers which elements they did (+1) or did not use (0) in each of the five lessons. A total of 35 elements were measured.
Process questionnaire for students
Process evaluation among students dealt with the extent to which the boosters were read (yes/no). The three magazines (boosters) were evaluated during the third post-test (T4).
Questionnaire for outcomes among students
The questionnaire measured age, gender, type of school, class, knowledge, attitudes, social norms, encountered pressure to smoke, perceived smoking behavior, intentions and smoking behaviors.
Knowledge about smoking was measured by 12 multiple choice questions with three answering categories: yes (+1)/no (0)/I don't know (0). The questions were summed to form one knowledge index (range 0–12), 0 indicating that all items was answered inappropriate, 12 indicating that all the questions were answered correctly. Attitudes were assessed on a seven-point or five-point scale by 13 beliefs (b) about smoking (e.g. `If I smoke I damage my health') and a person's evaluation (e) of these consequences (e.g. `Damaging my health is bad'). Beliefs and evaluations measured social (acceptance by friends, making contacts, being bullied less and sociable), personal (relieving boredom, relieving tension and tastes good) and health consequences of smoking (bad for your health, breathing problems, passive smoking, bad physical condition, coughing and unwise). The beliefs were multiplied by the evaluations and then summed to form one attitude scale (
b*e; Cronbach's 
= 0.78).
Social norms were measured by seven questions on a seven-point scale assessing the normative beliefs of the important persons (nb) and the corresponding motivations to comply with these opinions (mc). Normative beliefs measured to what degree, according to the students' opinion, important referents (such as their father) thought that they should or should not smoke. Motivations to comply items measured the willingness to meet the wishes of these referents (e.g. `I agree with what my friends think'). The normative beliefs and motivations to comply were multiplied and summed to form one social norm scale (nb*mc; Cronbach's = 0.76).
To sum, attitude and social norm questions were scored on a seven-point scale (–3 to +3). Only the questions concerning the motivation to comply and some (one-sided) attitudinal beliefs were measured on a five-point scale (0 to +3 or –3), all according to the measurement principles described by Ajzen and Fishbein (Ajzen and Fishbein, 1980).
Pressure to smoke was assessed by seven items on a five-point scale asking to what extent students felt pressure to smoke from several referents [range from `very often' (+4) to `never'(0)]. The questions were summed to form one social pressure index.
Perceived behavior measured the smoking behavior of six persons on a two-point scale [`smokes' (+1) or `does not smoke' (0)]. Also these items were summed to form one index.
The social norms, pressure and perceived behavior scales measured the behaviors and norms of: parents (father, mother), brothers/sisters, friends, best friend and family. The social norms and pressure scales also included cigarette industries.
Self-efficacy expectations (Cronbach's = 0.74) were assessed by seven questions on a seven-point scale assessing how confident a student was that he would be able to (1) refuse a cigarette offered to him, (2) to give a reason why he did not want a cigarette offered, (3) remain a non-smoker, (4) respond when he was bullied because he did not want to smoke, (5) not to smoke when his friends smoke, (6) say something about someone's smoking in a room when smoking was allowed and (7) say something about someone's smoking in a room when smoking was not allowed [range `yes, certainly' (+3) to `no, certainly not' (–3)].
Intention to smoke ( = 0.94) was measured by five items on a seven-point scale [from `certainly do' (+3), to `certainly do not' (–3) intend to smoke: in general, in discos, with friends, in my own room, with parents].
Smoking behavior was based on self-reports, categorizing a student as: (1) never a smoker: never smoked one (puff of a) cigarette; (2) initial smoker: tried smoking up to 5 times; (3) initial smoker: tried smoking more often, but is not a smoker now; (4) occasional smoker: smokes occasionally but not every week; (5) weekly smoker: smokes at least one cigarette per week; (6) daily smoker: smokes at least one cigarette per day.
Validation of self-reports was not feasible because of organizational and financial constraints. However, self-reports have been demonstrated to be accurate when anonymity and confidentiality were assured and an identification coding system was used (Akers et al., 1983; Hansen et al., 1985; Murray and Perry, 1987; Ransom, 1992; Abernathy and Bertrand, 1992; Botvin and Botvin, 1992; USDHHS, 1990). At each assessment, students provided an identification code consisting of the respondents initials, date of birth, gender, school and class. Apart from these, anonymity was assured.
Analysis
Program impact on smoking behavior was evaluated using logistic regression analyses. Since the program had a three-level sample (schools, classes and students), multilevel analyses using VARCL were performed (Longford, 1987, 1988; Bryk and Raudenbusch, 1992). VARCL, however, only gives an approximation to logistic regression, which sometimes runs into computational problems. Therefore, model reduction was applied using SPSS. The SPSS models were re-analyzed with VARCL to check if any substantial inter-school or inter-class effects occurred that might distort the SPSS results (De Vries et al., 1994). The VARCL analyses showed that less than 5% of the residual variance was due to between-class and between-school effects, and that the ß and SEs did not differ substantially between VARCL and SPSS; this was in agreement with the findings of an earlier Dutch study (De Vries et al., 1994). Therefore, it was decided to report the final SPSS models.
For the treatment effects on smoking, occasional, weekly and daily smokers were combined to form one group of smokers, while never smokers and initial smokers formed one group of non-smokers (0 = non-smokers, 1 = smokers). Post-treatment smoking was predicted from the school-level factors: treatment (T2: SI program, SIDM program and no program; T3 and T4: SI program with and without boosters, SIDM program with and without boosters, and no program), type of school (0 = 4-year track, 1 = 5/6-year track), class level (0 = grade 8, 1 = grade 9), and age, gender (0 = boys, 1 = girls), pre-treatment smoking and their interactions with treatment. Pre-treatment measures of attitude, social norms, pressure, perceived behavior, self-efficacy and intention were entered as covariates, since previous research had shown that these were predictors of smoking (De Vries and Kok, 1986; De Vries et al., 1994).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Descriptive findings
In the study boys and girls were almost equally represented. Table I
shows the increase of the percentage of smokers at T1, T2, T3 and T4 from pre-test to post-tests in smoking in the SI, SIDM and the control group. The smallest increase in smoking over time was found in the SI group with boosters, the control group reported the highest increase in smoking behavior.
|
Implementation
Programs
The core elements of the programs were used by the teachers (n = 59) as follows: on average 91% (n = 54) used the manuals, 90% (n = 53) used the video, activities were used by 84% (n = 50) of the teachers, 87% (n = 51) worked with peer-leaders and 91% (n = 54) used group activities. The least-implemented element (78%, n = 46) of the program were the summaries teachers were to hand out to students at the end of each lesson. Of the additional materials, the majority of the teachers (75%, n = 44) asked the students who decided not to smoke to write their name on a non-smoking poster and most of them (n = 39) hung the poster on the wall in the class. The majority of teachers (81%, n = 48) also handed out the quit-smoking brochures to the students, which students could give to smokers in their environment. A portion of the teachers (65%, n = 38) instructed the peer-leaders and most of them (n = 30) used the special training video.
Boosters
Of the students in the SIDM condition who received the three magazines (N = 339), 27% reported having read no magazine, 73% reported having read at least one magazine, 58% at least two magazines and 42% all three magazines. Of the students on the SI program who received the three magazines (N = 509), the respective figures were 41, 59, 45 and 36%.
Attrition
Attrition from pre-test (T1) to post-test 1 (T2) was 15.9% (N = 766). Logistic regression analyses with attrition as the dependent variable suggested that girls (OR = 0.74; 95% CI = 0.63–0.88), younger students (OR = 1.31; 95% CI = 1.16–1.48), students in the SIDM program (OR = 0.39; 95% CI = 0.31–0.48) and in the SI program (OR = 0.69; 95% CI = 0.56–0.85) compared with control group students and 4-year education students (OR = 1.21; 95% CI = 1.01–1.43) were less likely to drop out. Attrition at T1 was not related to pre-test smoking behavior.
Attrition from pre-test (T1) to post-test 2 (T3) was 24.3% (N = 1172). Girls (OR = 0.84; 95% CI = 0.73–0.97), younger students (OR = 1.32; 95% CI = 1.19–1.47), non-smokers (OR = 1.34; 95% CI = 1.20–1.49), second grade students (OR = 1.62; 95% CI = 1.33–1.96) and 4-year education students (OR = 1.33; 95% CI = 1.14–1.54) were less likely to drop out at the second post-test.
Attrition from pre-test (T1) to post-test 3 (T4) was 35.7% (N = 1722). Girls (OR = 0.85; 95% CI = 0.74–0.96), younger students (OR = 1.18; 95% CI = 1.07–1.31), non-smokers (OR = 1.38; 95% CI = 1.05–1.82), students in the control group compared with students in the SIDM program (OR = 1.57; 95% CI = 1.36–1.82), students in the SI program compared with control group students (OR = 0.61; 95% CI = 0.51–0.72), 4-year education students (OR = 1.22; 95% CI = 1.07–1.39) and students with a negative intention to start smoking (OR = 1.02; 95% CI = 1.004–1.03) were less likely to drop out at time of the third post-test.
In sum, the attitude analyses showed that at T2, T3 as well as T4 there were no significant interactions between pre-test smoking and treatment condition with respect to attrition.
Program effects on smoking at T2
At 6 months after the pre-test (T2), a significant interaction between treatment and pre-treatment smoking behavior was found. Therefore, separate analyses were necessary for pre-test non-smokers and pre-test smokers (Longford, 1987; Bryk and Raudenbusch, 1992). The results showed a significant treatment effect for pre-test among non-smokers of the SIDM group, both when comparing with the control group, as well as comparing the SI group with the control group (see Table II). No treatment effect was found for smokers.
|
Furthermore, students with a positive attitude and intention towards smoking, students with important people in their social environment who smoke (perceived behavior), and students who perceived pressure to smoke had an increased risk of being smokers at T2. No significant effects were found for the social norms, self-efficacy expectations, age, gender, class and type of school.
Program effects on smoking at T3
No significant interactions were found that suggested a different effect of the program for specific groups 12 months after the pre-test. The SI program with boosters had a significant better effect than the SI program without boosters and compared to the control group (Table III). The increase in smoking was 5.6, 12.2 and 12.6%, respectively. No significant effect was found for the SI program without boosters compared to the control group.
|
The SIDM program without boosters had a significantly greater significant impact than the control group (the increases were 8.1 and 12.6%, respectively), but the SIDM program with boosters did not differ in effectiveness compared to that program without boosters (with an increase in smoking of 10.6%).
The SI program with boosters was more effective in preventing smoking compared to the SIDM program with boosters, but the SI program without boosters was less effective than the SIDM program without boosters (see also Table III
and Table IV).
|
Older students, pre-test smokers, students with a positive attitude, and a positive intention, and students who perceived important smoking others had an increased risk of being a smoker at T3.
Program effects on smoking at T4
No differential effects of the programs for specific groups were found. The SI program with boosters differed significantly with the SI program without boosters and compared to the control group (Table IV), with increases of smoking of 9.7, 13.9 and 14.9%, respectively (Table I). The SI program without boosters did not differ with the control group.
No significant difference existed for the SIDM program with boosters compared to the program without boosters, while an almost significant effect (P < 0.07) was found for the SIDM program without boosters compared to the control group. The increases of smoking were 12.8, 10.4 and 14.9%, respectively.
Pre-treatment smokers, students with a positive attitude and a positive intention towards smoking, and students who perceived important smoking others and who experienced pressure to smoke had an increased risk of being a smoker at T4.
Intention to treat analysis
Since treatment was related with attrition at both T2 and T4, `intention to treat' analyses (Bulpitt, 1983) were conducted. All students were included and missing data were substituted by the last recorded smoking status. These analyses resulted in ßs, SEs and P values similar to those of the analyses excluding dropouts.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The findings presented in this study show that the SI approach was effective with respect to the reduction of the onset of smoking. At short term (6 months after pre-test) both programs were effective for the non-smokers. At 12 months after the pre-test the SI program was only effective when boosters were included, but the SIDM without boosters was still effective. At 18 months follow-up, only the SI program with boosters remained effective. Although many projects exist with reported evidence of success regarding smoking prevention, the comparison of effects among studies is complicated by many things; to mention the fact that some studies calculate their effectiveness on the basis of the number of non-smokers at pre-test, others compare smoking incidence figures at pre-test and post-test, some studies use regular smoking as the dependent variable, others use monthly smoking, etc. Although it is clear that interpretation and comparison of effectiveness of smoking prevention studies need to be done with great care, we regard our findings similar to other studies in this area, considering that in general the difference between treatment and non-treatment groups ranges from 25 to 60% and persists for more than 1 year (USDHHS, 1994). When comparing results to the more recent controlled, large-scale studies, the strength of this study is that numerous units of analysis were assigned to each experimental condition (in fact, 52 schools), it used a large sample size to explore the classroom and school characteristics (VARCL analysis).
The SI and the SIDM program had short-term preventive effects for non-smokers. No short-term treatment effects were found for smokers. The differential effects for non-smokers and smokers may be caused by the fact that both programs favored non-smoking, which, at first, may have evoked resistance in smokers. However, at T3 and T4 these differential effects were no longer found.
The SI program with boosters had a preventive effect on smoking at T3 and T4. The SI program without boosters, however, failed as expected to show a more enduring impact, which may imply that, indeed, these kinds of five lessons programs need to be followed by additional interventions to achieve longer lasting results.
The decision-making (SIDM) program without boosters was effective in preventing smoking compared to the control group at T3. At T4 this effect was smaller and had almost fallen to below statistical significance. Further, for the SIDM program the boosters power appeared to be lacking to indicate significant differences. The fact that the effect was much smaller than anticipated may have been caused by the fact that the boosters were the same for both the SI and the SIDM research conditions. This means that the boosters were not developed according to the decision-making stages by Janis and Mann (Janis and Mann, 1978), which may have caused some confusion among students, resulting in the above described outcome.
The process evaluations of the boosters in this study revealed that only small percentages of students of both programs reported having read all three magazines. Therefore, one might expect that the additional effect of boosters might increase if compliance with the booster intervention can be improved.
Although a relation was found between treatment and attrition at T2 and at T4, the students were not informed about the timing of assessments, and attrition could be attributed to factors such as illness and changing to another school. Moreover, intention to treat analyses including the drop-outs showed results comparable to those without this group. Consequently, the assumption can be made that the preventive effects of both programs at T2 and T4 are reliable.
A limitation of this study may have been that the self-reports on smoking behavior were not validated. However, self-reports have been demonstrated to be accurate when anonymity and confidentiality are assured, and an identification coding system is used (Akers et al., 1983; Hansen et al., 1985; Murray and Perry, 1987; USDHHS, 1990; Abernathy and Bertrand, 1992; Botvin and Botvin, 1992; Ransom, 1992). Another limitation is that no further long-term follow-ups could be implemented, precluding conclusions on the longer-term effectiveness of the programs. The `multiple testing' system used may have led to an increased risk of type I errors. However, using Bonferroni corrections did not substantially change the conclusions, since most P values were either >0.05 or <0.02, and trends were consistent between T3 and T4.
Finally, both programs can be used to prevent smoking in grades 8 and 9 of high schools. If the SI program is used, the booster should be included. Further research should determine whether boosters with decision-making related issues, e.g. issues based on the five steps to make a decision, contribute to the longer-term preventive effects of the decision-making program. Additional research is also needed to assess the long-term effects of the SI program with boosters, and to examine if a more intensive use of them is necessary and even more effective. Further research is needed also to assess whether regular distribution of further magazines can contribute to long-term effects of the SI program. In this respect, the interval between distributions, and the question up to what grade or age the magazines should be distributed, should be assessed as well.
|
Notes |
|---|
1 In the Netherlands, adolescents with different academic capabilities follow different educational tracks, with a duration of 4 years for people with the lowest academic capabilities and a maximum of 6 years for students with the highest capabilities. The latter group includes students that are preparing for university.
2 District health centers have the responsibility for public health care and health protection. They carry out activities to analyze the health status of the community, to combat infective diseases, and to prevent diseases in youth and other risk groups.
|
Acknowledgments |
|---|
This study was supported by a grant from the Dutch Cancer Foundation. The contribution of H. De V. was made possible because of an 8-year funding by the Dutch Cancer Society. We wish to thank all schools and health educators that contributed to the project.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Abernathy, T. J. and Bertrand, L. D. (1992) Preventing cigarette smoking among children: results of a four-year evaluation of the PAL program. Canadian Journal of Public Health, 83, 226–229.[Web of Science][Medline]
Ajzen, I. and Fishbein, M. (1980) Understanding Attitudes and Predicting Social Behavior. Prentice-Hall, Englewood Cliffs, NJ.
Akers, R. L., Massey, J., Clarke, W. and Lauer, R. M. (1983) Are self-reports of adolescent deviance valid? Biochemical measures, randomized response, and the bogus pipeline in smoking behavior. Social Forces, 62, 234–251.
Arkin, R. M., Roemhild, H. F., Johnson, L. A., Luepker, R. V. and Murray, D. M. (1981) The Minnesota Smoking Program. The Journal of School Health, 11, 611–616.
Bandura, A. (1986) Social Foundations of Thought and Action: A Social Cognitive Theory. Prentice-Hall, Englewood Cliffs, NJ.
Botvin, G. J., Baker, E., Filazzola, A. D. and Botvin, E. M. (1990) A cognitive-behavioral approach to substance abuse prevention: one-year follow-up. Addictive Behaviors, 15, 47–63.[Web of Science][Medline]
Botvin, G. J. and Botvin, E. M. (1992) Adolescent tobacco, alcohol, and drug abuse: prevention strategies, empirical findings, and assessment issues. Developmental Behavioral Pediatrics, 13, 290–301.
Bryk, A. S. and Raudenbusch, S. W. (1992) Hierarchical Linear Models: Applications and Data Analysis Methods. Sage, Newbury Park, CA.
Chatrou, M. (1992) Determinants of smoking and smoking prevention in Dutch adolescents. Dissertation. DSWO Press, Leiden University, Leiden.
De Vries, H. (1989a) Towards primary cancer prevention: the Dutch ABC framework. European Journal for Cancer and Clinical Oncology, 25, 1025.
De Vries, H. (1989b) Smoking prevention in Dutch adolescents. Dissertation. Datawyse, Maastricht.
De Vries, H. and Kok, G. J. (1986) From determinants of smoking behaviour to the implications for a prevention programme. Health Education Research, 1, 85–94.
De Vries, H., Weyts, W., Dijkstra, M. and Kok, G. J. (1992) The utilization of qualitative and quantitative data for health education program planning, implementation and evaluation: a spiral approach. Health Education Quarterly, 19, 101–115.[Web of Science][Medline]
De Vries, H., Backbier, E., Dijkstra, M., van Breukelen, G., Parcel, G. and Kok, G. J. (1994) A Dutch social influence smoking prevention approach for vocational school students. Health Education Research, 9, 365–374.
Dijkstra, M. (1995) Development, implementation and evaluation of two Dutch smoking prevention programs. Dissertation. Datawyse/UPM, Maastricht.
Dijkstra, M., De Vries, H. and Parcel, G. (1993) The linkage approach applied to a school-based smoking prevention program in the Netherlands. The Journal of School Health, 63, 339–342.[Web of Science][Medline]
Elder, J. P., Wildey, M., de-Moor, C., Sallis, J., Jr., Eckhardt, L., Edwards, C., Erickson, A., Golbeck, A., Hovell, M., Johnston, D., et al. (1993) The long-term prevention of tobacco use among junior high school students: classroom and telephone interventions. American Journal of Public Health, 83, 1239–1244.
Ellickson, P. L. and Bell, R. M. (1990) Drug prevention in junior high: a multi-site longitudinal test. Science, 247, 1299–1305.
Evans, R. I., Rozelle, R. M., Mittelmark, M. B., Hansen, W. B., Bane, A. L. and Havis, J. (1978) Deterring the onset of smoking in children: knowledge of immediate physiological effects and coping with peer pressure, media pressure and parent modeling. Journal of Applied Social Psychology, 8, 126–135.
Glynn, T. J. (1989) Essential elements of school-based smoking prevention programs. The Journal of School Health, 59, 181–188.[Web of Science][Medline]
Hansen, W. B., Malotte, C. K. and Fielding, J. E. (1985) The bogus pipeline revisited: the use of the threat of detection as a means of increasing self-reports of tobacco use. Journal of Applied Psychology, 70, 789–792.[Web of Science][Medline]
Janis, I. L. and Mann, L. (1978) Decision Making: A Psychological Analysis of Conflict, Choice, and Commitment. Free Press, New York.
Klepp, K. I., Halper, A. and Perry, C. L. (1986) The efficacy of peer leaders in drug abuse prevention. Journal of School Health, 56, 407–411.
Longford, N. T. (1987) A fast scoring algorithm for maximum likelihood estimation in unbalanced mixed models with nested random effects. Biometrika, 74, 817–827.
Longford, N. T. (1988) VARCL: software for variance component analysis of data with hierarchically nested random effects maximum likelihood. Manual. Educational Testing Service, Pinceton, NJ.
Mudde, A. N. (1994) The development and evaluation of a community and a mass media approach to smoking cessation. Dissertation. Maastricht.
Murray, D. M. and Perry, C. L. (1987) The measurement of substance use among adolescents: when is the `bogus pipeline' method needed? Addictive Behaviors, 12, 225–233.[Web of Science][Medline]
Murray, D. M., Pirie, P., Luepker, R. V. and Pallonen, U. (1989) Five- and six-year follow-up results from four seventh-grade smoking prevention strategies. Journal of Behavioral Medicine, 12, 207–218.[Web of Science][Medline]
Perry, C. L., Telch, M. J., Killen, M. J., Burke, J. and Maccaby, N. (1983) High school smoking prevention: the relative efficacy of varied treatments and instructions. Adolescence, 18, 561–566.[Web of Science][Medline]
Perry, C. L., Murray, D. M. and Klepp, K. I. (1987) Predictors of adolescent smoking and implications for prevention. Morbidity and Mortality Weekly Report, 36, 415–455.
Ransom, M. V. (1992) Peer perceptions of adolescent health behaviors. Journal of School Health, 62, 238–242.
US Department of Health and Human Services (1990) The Health Benefits of Smoking Cessation: A Report of the Surgeon General. Public Health Service. DHHS publ. no. CDC 90-8416. USDHHS, Atlanta, CA.
US Department of Health and Human Services (1994) Preventing Tobacco Use among Young People: A Report of the Surgeon General. USDHHS, Atlanta, GA.
Received on November 16, 1997; accepted on November 3, 1998
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Ausems, I. Mesters, G. van Breukelen, and H. De Vries Smoking among Dutch elementary schoolchildren: gender-specific predictors Health Educ. Res., October 1, 2009; 24(5): 818 - 828. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Smeets, J. Brug, and H. de Vries Effects of tailoring health messages on physical activity Health Educ. Res., June 1, 2008; 23(3): 402 - 413. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. C. Vijgen, P. H. M. van Baal, R. T. Hoogenveen, G. A. de Wit, and T. L. Feenstra Cost-effectiveness analyses of health promotion programs: a case study of smoking prevention and cessation among Dutch students Health Educ. Res., April 1, 2008; 23(2): 310 - 318. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. B. Buller, R. Borland, W. G. Woodall, J. R. Hall, J. M. Hines, P. Burris-Woodall, G. R. Cutter, C. Miller, J. Balmford, R. Starling, et al. Randomized Trials on Consider This, a Tailored, Internet-Delivered Smoking Prevention Program for Adolescents Health Educ Behav, April 1, 2008; 35(2): 260 - 281. [Abstract] [PDF]
|
|
|
|
 |
|
 S. Panday, S. P. Reddy, R. A. C. Ruiter, E. Bergstrom, and H. de Vries Determinants of smoking among adolescents in the Southern Cape-Karoo region, South Africa Health Promot. Int., September 1, 2007; 22(3): 207 - 217. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Vartiainen, M. Pennanen, A. Haukkala, F. Dijk, R. Lehtovuori, and H. De Vries The effects of a three-year smoking prevention programme in secondary schools in Helsinki Eur J Public Health, June 1, 2007; 17(3): 249 - 256. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. de Vries, F. Dijk, J. Wetzels, A. Mudde, S. Kremers, C. Ariza, P. D. Vitoria, A. Fielder, K. Holm, K. Janssen, et al. The European Smoking prevention Framework Approach (ESFA): effects after 24 and 30 months Health Educ. Res., February 1, 2006; 21(1): 116 - 132. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Panday, S. P. Reddy, R. A. C. Ruiter, E. Bergstrom, and H. de Vries Determinants of smoking cessation among adolescents in South Africa Health Educ. Res., October 1, 2005; 20(5): 586 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. T. H. M. Maassen, S. P. J. Kremers, A. N. Mudde, and B. M. Joof Smoking initiation among Gambian adolescents: social cognitive influences and the effect of cigarette sampling Health Educ. Res., October 1, 2004; 19(5): 551 - 560. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Hanewinkel and M. Asshauer Fifteen-month follow-up results of a school-based life-skills approach to smoking prevention Health Educ. Res., April 1, 2004; 19(2): 125 - 137. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. A. Foshee, K. E. Bauman, S. T. Ennett, G. F. Linder, T. Benefield, and C. Suchindran Assessing the Long-Term Effects of the Safe Dates Program and a Booster in Preventing and Reducing Adolescent Dating Violence Victimization and Perpetration Am J Public Health, April 1, 2004; 94(4): 619 - 624. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Markham, P. Aveyard, H. Thomas, A. Charlton, M. L. Lopez, and H. De Vries What determines future smoking intentions of 12- to 13-year-old UK African-Caribbean, Indian, Pakistani and white young people? Health Educ. Res., February 1, 2004; 19(1): 15 - 28. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Ausems, I. Mesters, G. van Breukelen, and H. De Vries Effects of in-school and tailored out-of-school smoking prevention among Dutch vocational school students Health Educ. Res., February 1, 2004; 19(1): 51 - 63. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. de Vries, A. Mudde, S. Kremers, J. Wetzels, E. Uiters, C. Ariza, P. D. Vitoria, A. Fielder, K. Holm, K. Janssen, et al. The European Smoking Prevention Framework Approach (ESFA): short-term effects Health Educ. Res., December 1, 2003; 18(6): 649 - 663. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. McBride A systematic review of school drug education Health Educ. Res., December 1, 2003; 18(6): 729 - 742. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. de Vries, A. Mudde, I. Leijs, A. Charlton, E. Vartiainen, G. Buijs, M. P. Clemente, H. Storm, A. G. Navarro, M. Nebot, et al. The European Smoking prevention Framework Approach (EFSA): an example of integral prevention Health Educ. Res., October 1, 2003; 18(5): 611 - 626. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. de Vries, R. Engels, S. Kremers, J. Wetzels, and A. Mudde Parents' and friends' smoking status as predictors of smoking onset: findings from six European countries Health Educ. Res., October 1, 2003; 18(5): 627 - 636. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M R Crone, S A Reijneveld, M C Willemsen, F J M van Leerdam, R D Spruijt, and R A H. Sing Prevention of smoking in adolescents with lower education: a school based intervention study J Epidemiol Community Health, September 1, 2003; 57(9): 675 - 680. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||