This art icle was downloaded by: [ Universit y of Texas] On: 05 Sept em ber 2012, At : 10: 22 Publisher: Taylor & Francis I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK Journal of Health Communication: International Perspectives Publicat ion det ails, including inst ruct ions f or aut hors and subscript ion inf ormat ion: ht t p: / / www. t andf online. com/ loi/ uhcm20 Media and Community Campaign Effects on Adult Tobacco Use in Texas a ALFRED McALISTER , THEODORE C. MORRISON a , ANGELA F. MESHACK , AMELIE RAMIREZ c VANCE RABIUS & PHILIP HUANG a b a , SHAOHUA HU a b , , KIPLING GALLION d Universit y of Texas School of Public Healt h, Houst on, Texas b Baylor College of Medicine, San Ant onio, Texas c American Cancer Societ y, Aust in, Texas d Texas Depart ment of Healt h, Aust in, Texas Version of record f irst published: 17 Aug 2010 To cite this article: ALFRED McALISTER, THEODORE C. MORRISON, SHAOHUA HU, ANGELA F. MESHACK, AMELIE RAMIREZ, KIPLING GALLION, VANCE RABIUS & PHILIP HUANG (2004): Media and Communit y Campaign Ef f ect s on Adult Tobacco Use in Texas, Journal of Healt h Communicat ion: Int ernat ional Perspect ives, 9: 2, 95-109 To link to this article: ht t p: / / dx. doi. org/ 10. 1080/ 10810730490425231 PLEASE SCROLL DOWN FOR ARTI CLE Full t erm s and condit ions of use: ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. The publisher does not give any warrant y express or im plied or m ake any represent at ion t hat t he cont ent s will be com plet e or accurat e or up t o dat e. The accuracy of any inst ruct ions, form ulae, and drug doses should be independent ly verified wit h prim ary sources. The publisher shall not be liable for any loss, act ions, claim s, proceedings, dem and, or cost s or dam ages what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h or arising out of t he use of t his m at erial. Journal of Health Communication, Volume 9: 95–109, 2004 Copyright # Taylor & Francis Inc. ISSN: 1081-0730 print/1087-0415 online DOI: 10.1080/10810730490425231 Media and Community Campaign Effects on Adult Tobacco Use in Texas Downloaded by [University of Texas] at 10:22 05 September 2012 ALFRED MCALISTER THEODORE C. MORRISON SHAOHUA HU ANGELA F. MESHACK University of Texas School of Public Health, Houston, Texas AMELIE RAMIREZ KIPLING GALLION Baylor College of Medicine, San Antonio, Texas VANCE RABIUS American Cancer Society, Austin, Texas PHILIP HUANG Texas Department of Health, Austin, Texas The present study reports on the effects on adult tobacco cessation of a comprehensive tobacco-use prevention and cessation program in the state of Texas. Differences in cessation rates across treatment conditions were measured by following a panel of 622 daily smokers, recruited from the original cross-sectional sample, from baseline to follow-up. The adult media campaign combined television, radio, newspaper and billboard advertisements featuring messages and outreach programs to help adults avoid or quit using tobacco products. The ads also promoted quitting assistance programs from the American Cancer Society Smokers’ Quitline, a telephone counseling service. The cessation component of the intervention focused on increasing availability of and access to cessation counseling services and pharmacological therapy to reduce nicotine dependence. Both clinical and community-based cessation programs were offered. Treatment areas which combined cessation activities with high level media campaigns had a rate of smoking reduction that almost tripled rates in areas which received no services, and almost doubled rates in areas with media campaigns alone. Analyses of the dose of exposure to media messages about smoking cessation show greater exposure to television and radio messages in the areas where This research was supported in part by grant number 202-6-0802 from the National Cancer Institute. Address correspondence to Alfred McAlister, Center for Health Promotion and Prevention Research, University of Texas Houston School of Public Health, P.O. Box 20036, Houston, TX 77225, USA. E-mail: amcalister@sph.uth.tmc.edu 95 96 A. McAlister et al. high level media was combined with community cessation activities than in the other areas. Results also show that exposure to media messages was related to processes of change in smoking cessation and that those processes were related to the quitting that was observed in the group receiving the most intensive campaigns. Downloaded by [University of Texas] at 10:22 05 September 2012 Introduction The state of Texas won a lawsuit against the tobacco industry for compensation of health costs related to tobacco use and the Texas legislature has allocated some funds for efforts to reduce tobacco use. To guide legislative decision-making with respect to spending for this purpose, funds from the settlement (approximately $9 million per year) were allocated by the legislature to the Texas Department of Health for the implementation and evaluation of pilot activities in selected areas of Texas. This report presents the result of a short-term, quasi-experimental assessment of media and community campaigns to promote smoking cessation among adults. Studies of comprehensive programs at the community and regional levels have entertained long-term success in reducing tobacco use. The North Karelia Project in Finland showed reduction in adult and youth tobacco use at ten to fifteen year follow-up. Its program components included cessation and prevention activities involving mass media, public policy, schools, health care providers, community organizations and both formal and informal opinion leaders (McAlister, Puska, Koskela, Pallonen, & Maccoby, 1980; Puska, Salonen, Tuomilento, & Koskela, 1982). In South Texas, a four-year community study using multi-component media and community programs achieved significant increases in maintained smoking cessation among adults (McAlister et al., 1992). Other programs such as the Stanford Five-City Study in California, the Minnesota Heart Health Study, and a study of combined school and media activities in Vermont have demonstrated the effectiveness of comprehensive programs on tobacco use among children and adults (Farquhar, Fortmann, Flora, Taylor, Haskell, & Williams, et al., 1995; Perry, Kelder, Murray, & Klepp, 1992; Flynn et al., 1994). In contrast, programs using only a single component such as cessation (e.g., COMMIT, 1995) or prevention in schools (e.g., Ellickson, Bell, & McGuigan, 1993), have not achieved significant longterm effects on adult or youth tobacco use (US DHHS, 2000). Since multi-component approaches have shown marked reductions in tobacco use and increases in cessation in contrast to single-component designs, federal tobacco control experts now recommend a ‘‘comprehensive’’ approach with combined prevention and cessation activities including mass media, school-based programs, health care providers, and outreach to community organizations US (DHHS, 1999). The present study reports on the effects on adult tobacco prevalence and cessation of a comprehensive tobacco-use prevention and cessation program in the state of Texas. It was hypothesized that the areas that received an intensive media campaign combined with a community cessation program would exhibit the highest rates of cessation at follow-up when compared with areas that received no programs or less intensive campaigns. It was further hypothesized that areas which received intensive media campaigns combined with a comprehensive cessation and prevention program would show the lowest tobacco-use prevalence at follow-up. Methods The original study was designed to test combinations of media levels (none, low-level media, or high-level media) and program options (no program, law enforcement pro- Downloaded by [University of Texas] at 10:22 05 September 2012 Campaign Effects on Adult Tobacco Use in Texas 97 grams only, cessation programs only, school-community prevention programs only, or all preceding programs combined comprehensively) on tobacco-use prevalence and cessation. There were a total of 14 possible combinations since the comprehensive program was not combined with a no-media campaign. Each of 19 areas in eastern Texas was assigned to one of the 14 treatment conditions (Figure 1). Harris County, which includes the city of Houston, was the only major metropolitan area in the study, and due to it’s size and the fact that it comprised one media market, it was assigned to low-level media and broken into five areas that received each of the program conditions (Figure 2). All other treatment conditions were delineated by county. Eastern Texas was chosen as the overall study area due to its higher rates of smoking and smoking related illnesses. All treatment areas had populations of over 100,000. Assignment to condition was constrained by overlap of media markets. Areas with the greatest racial/ethnic diversity were assigned to the comprehensive treatment condition. Cessation programs alone were offered in four treatment areas and in three treatment areas as part of the comprehensive program. This study tests the effects on smoking cessation of combinations of media levels with or without programs including a cessation component (alone or as part of the comprehensive program) among a sub-sample of daily smokers. The original study was designed by a consortium of researchers at the University of Texas in Austin, the University of Texas Health Science Center at Houston School of Public Health, and the Texas Department of Health. Data collection and implementation of program activities were done by independent contractors not related to each other. These independent contractors included survey research centers, media consultant firms, and health educators. Data analysis was done by research associates and a professor at the University of Texas Health Science Center at Houston School of Public Health. Panel Survey and Cross-Sectional Surveys A quasi-experimental cross-sectional study design was used to test for significant differences of program effects on tobacco-use prevalence and tobacco-use cessation across treatment conditions. To determine program effects on tobacco-use prevalence, two independent random samples of 9,407 (baseline) and 8,974 (follow-up) adults were interviewed by telephone at baseline (April–May 2000) and seven-month follow-up (November-December 2000). The independent cross-sectional samples at both baseline and follow-up contained individuals from across the state of Texas who did not reside in the treatment areas (N ¼ 999 at baseline, N ¼ 931 at follow-up). Differences in cessation rates across treatment conditions were measured by following a panel of 622 daily smokers, recruited from the original cross-sectional baseline sample, from baseline to seven-month follow-up. Survey Methods A random-digit-dialing sample was purchased from Survey Sampling, Inc. of Fairfield, Connecticut. They identified working telephone exchanges throughout the entire state of Texas, and systematically generated telephone numbers by four-digit randomization. Zip codes belonging to the nineteen study areas were identified based on three digit exchanges within area codes. In addition to these steps, Survey Sampling, Inc. compared the random-digit numbers against the 9.2 million Yellow Pages numbers in their database, eliminating any numbers that matched, so that the chances of reaching a residential number were increased. Since certain people are more likely to answer the phone than others, and certain people are more likely to be willing to participate in surveys than others, the respondent within each household with the most recent birthday was selected. Downloaded by [University of Texas] at 10:22 05 September 2012 98 FIGURE 1 Map of treatment conditions. Downloaded by [University of Texas] at 10:22 05 September 2012 FIGURE 2 Map of harris county treatment conditions. 99 Downloaded by [University of Texas] at 10:22 05 September 2012 100 A. McAlister et al. The data collection team for this project consisted of a field director who was in charge of constant monitoring of the sample, sample control, verification of interviews, scheduling of monitoring, and providing feedback to interviewers and supervisors. The core data collection team consisted of fifty interviewers and seven supervisors. All interviewers and supervisors received training including four hours of classroom instruction, computer tutorials, and practice interviewing sessions before working on the project. The data collection team was familiarized with the overall study objectives, interviewing techniques, the use of the Computer Assisted Telephone Interviewing system, the survey instrument, the definitions of key terms specific to the study, and the University of Texas statement of interviewer ethics and obligations. Interviewer performance was evaluated throughout the data collection period. A monitoring team used a standardized evaluation instrument to randomly assess individual interviewing performance in the use of appropriate feedback, reading verbatim, proper speech and pronunciation, interviewing pace, and general rapport with respondents. Interviewers were monitored several times during the study period and received feedback on their performance. Survey supervisors also verified ten percent of the completed interviews during the data collection period to ensure that interviews were conducted with the correct respondent. The verification procedure consisted of re-contacting respondents of completed surveys and asking them whether they participated in the survey. Potential respondents were told: ‘‘We are calling on behalf the Texas Department of Health. We’re doing a study of tobacco use among Texas residents. Your phone number has been chosen randomly to be included in the study, and we’d like to ask some questions about you and your use of tobacco.’’ Potential respondents had to be at least 18 to participate. Spanish-speaking interviewers were available to interview respondents who spoke only Spanish. Measures The interview contained one hundred and four items. Not all respondents were asked all questions and assessment of many items was dependent upon the respondents’ answers to previous questions. Measures included items on demographics, tobacco use, cessation, intentions, knowledge, attitudes, beliefs, brand preferences, drinking behaviors, depression, relaxation, exposure to media messages for and against tobacco use, and program activities. Respondents to the baseline survey were asked ‘‘Do you now smoke cigarettes everyday, some days, or not at all?’’ Those who answered ‘‘everyday’’ to this question and agreed to be contacted at a later date are included in the current analysis. Questions on tobacco use came from the Centers for Disease Control’s Behavioral Risk Factor Surveillance System. The reliability and validity of these measures are supported in the literature (Bowlin, Morrill, Nafziger, Jenkins, Lewis, & Pearson, 1993; Nelson, Holtzman, Bolen, Stanwyek, & Mack, 2001). Exposure to the media campaign was measured by three questions assessing frequency of having been exposed to media messages through television, radio, and newspaper advertisements over the past 30 days. Answer choices were (1) never, (2) 1 to 3 times, (3) 1 to 3 times per week, (4) daily or almost daily, and (5) more than once a day. Processes of change variables adapted from Prochaska and DiClemente (1983) measured agreement with attitudes regarding smoking and occurrence of behaviors associated with smoking cessation. Respondents were asked to rate how much they agree or disagree with the following statements: ‘‘Smoking helps a person relax;’’ ‘‘smoking helps a person concentrate and do better work;’’ ‘‘a person’s smoking can affect the Campaign Effects on Adult Tobacco Use in Texas 101 health of those around them;’’ ‘‘most of the people close to me disapprove of smoking;’’ ‘‘I think about information from articles and advertisements on how to quit smoking;’’ ‘‘I get embarrassed that I have to smoke;’’ ‘‘I tell myself I can quit smoking if I want to;’’ ‘‘I have someone I can count on when I’m having problems with smoking;’’ ‘‘I do something else instead of smoking when I need to relax or deal with tension;’’ and ‘‘I put things around my home that remind me not to smoke.’’ Answer choices were: (1) strongly disagree, (2) disagree, (3) neither agree nor disagree, (4) agree, and (5) strongly agree. The reliability and validity of these measures are supported in the literature (Crittendan, Manfredi, Warnecke, Cho, & Parsons, 1998; O’Connor, Carbonari, & DiClemente, 1996; Rustin & Tate, 1993; Velicer, Norman, Fava, & Prochaska, 1999). Downloaded by [University of Texas] at 10:22 05 September 2012 Media Campaign for Cessation The adult media campaign combined television, radio, newspaper, and billboard advertisements. All of the media messages promoted quitting assistance programs such as the American Cancer Society Smokers’ Quitline, a telephone counseling service available to smokers in all of the study areas. Additionally, posters were created for dental and physician offices encouraging patients to talk to their health care providers about cessation options. One third of media-buy resources were used to target minority audiences. Ads were created in English, Spanish, and Vietnamese. The primary target audience was adults aged 25–49 who use tobacco, while a secondary target audience included younger adults who may not yet be aware of their nicotine addiction. Many animated television ads and radio spots were created that included the Texas Duck who is against smoking and these theory-based ads targeted adolescents’ attitudes about tobacco use gleaned from school surveys of over 30,000 students. While adults and adolescents were both exposed to these ads, their contribution to smoking cessation among adolescents was not ascertained because the panel of daily smokers did not include people under the age of 18. Media messages were developed with input from community forums, focus groups, and pre-testing. Messages were grounded in social learning theory (Bandura, 1986) and the transtheoretical model (Prochaska & DiClemente, 1983), and included concepts such as modeling, social reinforcement for behavior change, and emotional arousal. Television ads included two developed by the Centers for Disease Control (CDC). The first ad featured the brother of the Marlboro Man who laments his passing as a result of lung cancer. The second ad demonstrated how a man who wants to quit smoking keeps a picture of his daughter in his cigarette pack as a reminder of someone who cares about him and wants him to quit. Radio spots included versions of the two CDC ads. One radio spot produced for the current study featured children telling their parents that they are concerned about their smoking and they wonder ‘‘what will happen to us if you die?’’ This ad also appeared in print format in local newspapers. All ads promoted the American Cancer Society Smokers’ Quitline and the radio and print ads encouraged smokers to ask their doctor or pharmacist about tools to help them quit such as the nicotine patch. Additionally, a ten second public service radio announcement promoting the Quitline was broadcast during morning drive times. Community Programs for Cessation The cessation component of the intervention focused on increasing availability of and access to cessation counseling services and pharmacological therapy to reduce nicotine dependence. Both clinical and community-based cessation programs were offered. Health Downloaded by [University of Texas] at 10:22 05 September 2012 102 A. McAlister et al. educators made face-to-face contact with clinicians, including physicians, dentists, and pharmacists. Clinicians were provided with a tool kit and were educated and encouraged to identify tobacco users, encourage a quit attempt, and provide or refer to counseling and/or pharmacological therapy as needed. The tool kits included tobacco assessment sheets and tear-off patient information sheets to go into patients’ files and to be given to the tobacco-using patient by the clinician. Tool kits also included a laminated card giving point-by-point instructions on how to counsel a tobacco-using patient to make a quit attempt. Community-based programs included public education and environmental change strategies such as worksite smoking bans. Presentations were held in a variety of settings including worksites, churches, and day care centers. Education efforts were reinforced by media messages and buttressed by promotion of the American Cancer Society Smokers’ Quitline. One community-based effort was the Quit and Win contest sponsored by COMPROTax offices in Jefferson County, Texas. The Quit and Win contest lasted three weeks and provided a drawing for a $2000.00 savings bond for smokers who pledged to quit smoking. The contest generated extensive media coverage including a kick-off media conference held on the steps of City Hall in Port Arthur, sponsors, local coordinators and mayoral representatives being booked on radio and television talk shows, and daily public service announcements (PSA) advertising the Quitline and contest by sponsoring radio and television stations. Newspaper promotions included PSAs and editorials featuring role-model stories on efforts to quit. Churches and civic organizations provided their members with Quit and Win entry pledge forms. Some hosted testimonies from members about their previous smoking habits, quitting, and added value to their lives since they stopped smoking. Results The Samples Of the panel of 1,069 daily cigarette smokers selected from the baseline survey, 622 (58%) responded to the follow-up survey, with most attrition coming from invalid numbers (234). Of the remaining 835 valid numbers, 12 respondents and 16 informants refused to participate when contacted and 185 were not completed due to no-answer, busy, and answering machines. There were no significant differences between the treatment conditions in the response rate for the follow-up survey. The 622 daily-smokers included 62.9% females, 82.9% ‘‘White or Ango or European descent/non-Hispanic,’’ 8.3% ‘‘Black or African-American/non-Hispanic,’’ 6.0% ‘‘Hispanic or Mexican or Mexican-American or Latino,’’ 0.2% ‘‘Asian: Chinese, Japanese, Korean, Vietnamese, Filipino or other Southeast Asian group, Indian, Pakistani or other South Asian group,’’ and 2.6% ‘‘other racial/ethnic group.’’ Ages of the 622 daily smokers are grouped as follows: 18 to 24 (9.6%), 25 to 44 (47.0%), 45 to 64 (36.5%), and 65 or older (7.0%). Education levels are grouped as follows: grades 1 through 11 (12.4%), grade 12 or GED (32.6%), 1 to 3 years of college (36.0%), 4 or more years of college (18.6%), and with 0.4% not responding. Using chi-square analyses, the demographics of the respondents (622) and nonrespondents (447) to the follow-up survey of the panel of daily smokers were compared. Significant differences were found for sex, age, and race/ethnicity. Of females, 60.6% completed the follow-up survey as compared to a completion rate of 54.4% for males. Completion rates for age groups are as follows: 18 to 24 (34.9%), 25 to 44 (57.8%), 45 to 64 (67.8%), and 65 and older (75.4%). Downloaded by [University of Texas] at 10:22 05 September 2012 Campaign Effects on Adult Tobacco Use in Texas 103 Demographic characteristics of the two independent random samples of 9,407 (baseline) and 8,974 (follow-up) were compared with 2000 census data by survey areas (Texas not including those assigned to a condition and all areas assigned to a condition combined). Statistically significant differences were found for all six chi-square comparisons with overrepresentations of females, Anglos (White, Non-Hispanic), and those in the 45 to 64 and 65 and older age groups. Such differences are common among random-digit-dialed samples. Demographic characteristics of the two random samples at baseline and follow-up were compared with each other using Pearson’s chi-square test statistic. There were significant differences for sex and race/ethnicity; age was not statistically different. The baseline sample was 64.3% female and the follow-up sample was 56.8% female. The racial/ethnic make-up of the two samples (baseline/follow-up percentages) was: Anglo (White, Non-Hispanic) (70.2/71.4), Black or African-American (12.6/13.4), Hispanic (13.1/11.1), Asian (1.8/1.6), and other (2.3/2.5). Media and Program Effects on Cessation Among the Panel of Adult Daily Smokers Complete cessation of smoking was reported by approximately two percent of the original panel of daily smokers and the sample size at this rate was too small to be reliably analyzed. However, a larger proportion (5.2%) reported that they were no longer smoking every day, and this group provided a sufficient number for statistical comparisons. The proportion ceasing daily smoking was calculated both as a proportion of follow-up panel survey respondents and as a proportion of the entire original panel of daily smokers. The latter method is the most conservative way to estimate cessation rates, as it assumes that all non-respondents have continued to smoke. According to either criterion, a higher rate of daily-smoking cessation was seen in the high-media areas than in the low- or no-media areas. There was also a relatively higher daily-smoking cessation rate in the cessation program delivery area than the other areas. However, these differences in daily-smoking cessation rates are not statistically significant. To determine whether media campaign level and program delivery type might have had a significant interactive effect, we separated the different media and program groups into five groups ranging from no media/no community cessation activities to high-level media/community cessation activities as shown in Figure 3. The percent reduction was calculated both as a proportion of the panel of daily smokers who participated in the follow-up interview and as a proportion of all smokers who were enrolled in the study. Among these follow-up participants there was a strongly significant trend toward higher daily-smoking cessation rates in the more intensive treatment groups (linear association, chi square ¼ 6.58, df ¼ 1, p < 0.01). The highest rate of daily-smoking cessation, close to 14%, was in the area with high-level media campaigns and cessation service delivery. There was less change (11%) in areas with cessation services and a low-level media campaign. Media campaigns without cessation services yielded cessation rates of only approximately 8%, while a cessation rate of 5% was found among those with no campaigns or services. The same pattern was found when reductions were calculated as a proportion of the original study groups (chi square ¼ 5.55, df ¼ 1, p < 0.02), with cessation rates ranging from 3% to 8%. While there were no observed differences in sex and age between the five study conditions for the panel of smokers, the proportion of African-Americans was significantly higher within the high-level media/community cessation activities group A. McAlister et al. Downloaded by [University of Texas] at 10:22 05 September 2012 104 FIGURE 3 Smoking cessation by experimental area. (20.6% vs. between 4.4% and 6.9% in the other conditions). Logistic regression analyses were performed to determine if ethnicity had an effect on the above outcomes. When ethnicity was included in the logistic models, the five-level experimental grouping variable was still a significant predictor of quitting daily smoking and ethnicity was not significant. Campaign Effects on Adult Tobacco Use in Texas 105 Downloaded by [University of Texas] at 10:22 05 September 2012 Combined Media and Program Effects on Tobacco-Use Prevalence among Independent Cross-Sectional Survey Respondents Both baseline and follow-up independent cross-sectional surveys had a response rate of approximately 60%. Adjusting for differences in age, gender, and education level, the daily cigarette smoking rate was 15.7% at baseline and 17.5% at follow-up. However, while the prevalence of daily cigarette smoking increased in all other areas, the proportion decreased from 0.18 (N ¼ 998) to 0.17 (N ¼ 701) in the area with the most intensive media campaign and the comprehensive program. The reduction in that area is not statistically significant when analyzed separately. But it is significantly different (general linear model, analysis of time by group interaction, F ¼ 5.0, df ¼ 1, 5265, p < 0.03) from the relative increase, from 0.13 (N ¼ 1,798) to 0.17 (N ¼ 1,761), in the prevalence of daily smoking that was observed among adults in the statewide sample combined with the areas where no media campaigns or community programs were provided. Linkages Between Experimental Group, Campaign Exposure, and Cessation Processes among the Panel of Daily Smokers A series of analyses was performed to study the interaction between program delivery, dose of exposure to media messages, and processes of smoking cessation. First, association between change in exposure to media messages about quitting smoking or not using tobacco, and whether subjects were in the high media with cessation program condition versus all others was tested using one-way ANOVAs. Three exposure change scores were calculated by subtracting baseline from follow-up answers to television, radio and print media exposure questions. The group receiving high level media and community cessation programs had significantly greater change scores than the other groups combined for television (F552,1 ¼ 3.54) and radio exposure (F555,1 ¼ 10.10). Next, Pearson correlations between follow-up exposure and process variables were computed to assess the influence of media exposure on processes involved in smoking cessation. Significant correlations between follow-up exposure variables and process variables are shown in Table 1. T-tests were performed to test for mean differences in process variables between those who reported daily-smoking cessation and those who maintained daily smoking. The mean level of agreement (range 1 to 5) with the following statements was significantly higher among those who stopped daily smoking: ‘‘I have someone I can count on when I’m having problems with smoking’’ (mean ¼ 3.46 versus 3.17, t ¼
1.83, p < 0.04), ‘‘I do something else instead of smoking when I need to relax or deal with tension’’ (mean ¼ 3.80 versus 2.94, t ¼
6.43, p < 0.001), and ‘‘I put things around my home that remind me not to smoke’’ (mean ¼ 2.53 versus 2.16, t ¼
2.50, p < 0.01). These are the same process variables that were related to television and radio exposure. Among those who were still smoking at all at follow-up, 162 (27.3%) had made an attempt to quit smoking. T-tests were calculated to determine if the process variables that were associated with campaign exposure were also related to quit attempts. Quit attempters were significantly more likely to agree that ‘‘I tell myself I am able to quit smoking if I want to’’ (mean ¼ 3.13 versus 2.85, t ¼
2.46, p < 0.01), ‘‘I think about information from articles and advertisements on how to quit smoking’’ (mean ¼ 3.39 versus 3.03, t ¼
3.49, p < 0.001), and ‘‘I put things around my home that remind me not to smoke’’ (mean ¼ 2.31 versus 2.14, t ¼
2.16, p < 0.02). 106 A. McAlister et al. Downloaded by [University of Texas] at 10:22 05 September 2012 TABLE 1 Significant (p < 0.05) Pearson Correlations Between Exposure and Process Variables at Follow-up Process1 T.V.2(N) Radio2(N) I think about information from articles and advertisements on how to quit smoking I tell myself I am able to quit smoking if I want to I have someone I can count on when I’m having problems with smoking I do something else instead of smoking when I need to relax or deal with tension I put things around my home that remind me not to smoke .13 (574) .08 (582) N.S. (524) .12 (533) .14 (566) .09 (572) .11 (573) .08 (581) N.S. (578) .14 (586) 1 Process variables answer choices were (1) strongly disagree, (2) disagree, (3) neither agree nor disagree, (4) agree, and (5) strongly agree. 2 Frequency of having been exposed to media messages through television and radio over the past 30 days. Answer choices were (1) never, (2) 1 to 3 times, (3) 1 to 3 times per week, (4) daily or almost daily, and (5) more than once a day. Discussion Consistent with previous studies noted in the introduction (e.g., McAlister et al., 1980; McAlister et al., 1982; McAlister et al., 1992; Farquhar et al., 1990; Perry et al., 1992), the results of this study support the conclusion that reductions in cigarette smoking can be achieved through community-level campaigns that combine high level media campaigns with cessation programs or comprehensive programs including cessation and community activities. Treatment areas which combined cessation activities with high level media campaigns saw a rate of smoking reduction that almost tripled rates in areas which received no services, and almost doubled rates in areas with media campaigns alone. Similarly, the area with high-level media combined with all program components saw a reduction in tobacco-use prevalence when compared with areas with no media campaign or program components. Of course, findings from a quasi-experimental study should be interpreted with caution. Despite the precautions taken to choose comparable experimental communities, random differences and unmeasured variations among communities can introduce error into estimations of experimental effects. Further, the use of self-report measures and attrition among our panel of baseline smokers may have introduced bi as and lowered the statistical power necessary to detect experimental differences. However, chi-square analysis using the entire sample showed that those lost to follow-up were no more or less likely at baseline to report planning to quit smoking in the next six months, suggesting Downloaded by [University of Texas] at 10:22 05 September 2012 Campaign Effects on Adult Tobacco Use in Texas 107 that attrition may not have affected follow-up quit rates. A comparison of respondents and non-respondents to the follow-up survey of daily smokers with regards to sociodemographic variables showed significant differences for sex, age, and race/ethnicity. In addition, there were differences with regards to sex and race/ethnicity between the two independent random samples and 2000 Census data. Considering the potential impact of such variables on smoking cessation (Coambs, Li, & Kozlowski, 1992; Gilpin & Pierce, 2002; Picardi, Bertoldi, & Morosini, 2002; Ward et al., 2002), one should use caution when interpreting or making generalizations beyond the results of this study. Finally the use of six-month follow-up measures can only allow us to make interpretations regarding the relatively short-term effectiveness of such a tobacco cessation program. The inferences about media and community campaign effects that can be drawn from this study are greatly strengthened by the findings with regard to the mechanisms through which reductions in smoking were achieved. Analysis of the dose of exposure to media messages about smoking cessation shows greater increase in exposure to television and radio messages in the areas where high-level media was combined with community cessation activities than in the other areas. Results also show that exposure to media messages is related to processes of change in smoking cessation and that those processes are related to the quitting and attempts to quit that were observed in the group receiving the most intensive campaign. The messages in the media campaign were focused on promotion of the telephone counseling service (Smokers’ Quitline) provided by the American Cancer Society. This generated 1,577 calls from smokers who were randomized to receive either counseling or self-help materials, and a separate study showed that the counseling significantly increased cessation rates among younger smokers (Rabius McAlister, Geiger, Huang, & Todd, under review). However, the Quitline callers were only a very small fraction (less than one half of one percent) of the entire population of smokers in the study areas. Thus it is interesting to note that the media campaign evidently influenced cessation processes among smokers who did not themselves call to receive counseling. This suggests that promotion of telephone counseling may have a ‘‘ripple effect’’ that goes beyond the quitting and quit attempts that are attributable directly to the receipt of counseling. The findings from this project, and from related studies of tobacco use prevention campaigns for young people, were reported in a summary form to the Texas Legislature during the winter of 2001 (University of Texas, 2001). The report concluded that significant reductions in adult tobacco use can be achieved through a combination of intensive media and community campaigns. This conclusion was accepted by key decision-makers and, during the spring of 2001, the legislature appropriated an additional $5 million per year for the Texas Department of Health to conduct intensive media and community campaigns in the areas that had no campaigns or limited campaigns in this pilot study. If further studies show continuing reductions in tobacco use, a statewide campaign will be considered. References Bandura, A. (1986). Social Foundations of Thought and Action. Englewood Cliffs, N.J.: PrenticeHall. Bowlin, S.J., Morrill, B.D., Nafziger, A.N., Jenkins, P.L., Lewis, C. & Pearson, T.A. (1993). Validity of cardiovascular disease risk factors assessed by telephone survey: the Behavioral Risk Factor Survey. Journal of Clinical Epidemiology, 46, 561–571. Coambs, R.B., Li, S. & Kozlowski, L.T. (1992). Age interacts with heaviness of smoking in predicting success in cessation of smoking. American Journal of Epidemiology, 135, 240–6. Downloaded by [University of Texas] at 10:22 05 September 2012 108 A. McAlister et al. COMMIT Research Group. (1995). Community Intervention Trial for Smoking Cessation (COMMIT): I. Cohort results from a four-year community intervention. American Journal of Public Health, 85, 183–192. Crittenden, K.S., Manfredi, C., Warnecke, R.B., Cho, Y.I. & Parsons, J.A. (1998). Measuring readiness and motivation to quit smoking among women in public health clinics: predictive validity. Addictive Behaviors, 23, 191–199. Ellickson, B.L., Bell, R.M. & McGuigan, K. (1993). Preventing adolescent drug use: long-term results of a junior high program. American Journal of Public Health, 83, 856–861. Farquhar, J.W., Fortmann, S.P., Flora, J.A., Taylor, C.B., Haskell, W.L., Williams, P.T., et al. (1990). Effects of community-wide education on cardiovascular disease risk factors. Journal of the American Medical Association, 264, 359–365. Flynn, B.S., Worden, J.K., Secker-Walker, R.H., Pirie, P.L., Badger, G.J., Carpenter, J.H., et al. (1994). Mass media and school interventions for cigarette smoking prevention. American Journal of Public Health, 84, 1148–1150. Gilpin, E.A. & Pierce, J.P. (2002). Demographic differences in patterns in the incidence of smoking cessation: United States 1950–1990. Annals of Epidemiology, 12, 141–50. McAlister, A., Puska, P., Koskela, K., Pallonen, U. & Maccoby, N. (1980). Mass communication and community organization for public health education. American Psychologist, 35, 375–39. McAlister, A., Puska, P., Salonen, J., Tuomilehto, J. & Koskela, K. (1982). Theory and action for health promotion: Illustrations from the North Karelia project. American Journal of Public Health, 72, 43–50. McAlister, A., Ramirez, A., Amezcua, C., Pulley, L., Stern, M. & Mercado, S. (1992). Smoking cessation in Texas-Mexico border communities: A quasi-experimental panel study. American Journal of Health Promotion, 6, 274–279. Nelson, D.E., Holtzman, D., Bolen, J., Stanwyck, C.A. & Mack, K.A. (2001). Reliability and validity of measures from the Behavioral Risk Factor Surveillance System. Sozial- und Praventivmedizin, 46 Suppl 1, S3-S42. O’Connor, E.A., Carbonari, J.P. & DiClemente, C.C. (1996). Gender and smoking cessation: a factor structure comparison of processes of change. Journal of Consulting & Clinical Psychology, 64, 130–138. Perry, C.L., Kelder, S.H., Murray, D.M. & Klepp, K. (1992). Community-wide smoking prevention: Long-term outcomes of the Minnesota Heart Health Program and the Class of 1989 Study. American Journal of Public Health, 9, 1210–1216. Picardi, A., Bertoldi, S. & Morosini, P. (2002). Association between the engagement of relatives in a behavioural group intervention for smoking cessation and higher quit rates at 6-, 12- and 24-month follow-ups. European Addiction Research, 8, 109–17. Prochaska, J.O. & DiClemente, C.C. (1983). Stages and processes of self-change of smoking: Toward an integrative model of change. Journal of Consulting & Clinical Psychology, 51, 390–395. Rabius, V., McAlister, A., Geiger, G., Huang, P. & Todd, R. Telephone Counseling Increases Cessation Rates among Young Adult Smokers. Under review. Rustin, T.A. & Tate, J.C. (1993). Measuring the stages of change in cigarette smokers. Journal of Substance Abuse Treatment, 10, 209–220. University of Texas Health Science Center at Houston. (2001). Texas Tobacco Prevention Initiative Media Campaign and Community Program Effects among Children and Adults. US Department of Health and Human Services. (1999). Best Practices for Comprehensive Tobacco Control Programs. Atlanta, Georgia: US-DHHS, Centers for Disease Control and Prevention, National Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, August. Campaign Effects on Adult Tobacco Use in Texas 109 Downloaded by [University of Texas] at 10:22 05 September 2012 US Department of Health and Human Services. (2000). Reducing Tobacco Use: A Report of the Surgeon General. Atlanta, Georgia: US-DHHS, Centers for Disease Control and Prevention, National Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. Velicer, W.F., Norman, G.J., Fava, J.L. & Prochaska, J.O. (1999). Testing 40 predictions from the transtheoretical model. Addictive Behaviors, 24, 455–469.