Health Education Research, Vol. 15, No. 3, 353-366,
June 2000
© 2000 Oxford University Press
Durability of tobacco control efforts in the 22 Community Intervention Trial for Smoking Cessation (COMMIT) communities 2 years after the end of intervention
Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, MP-702, PO Box 19024, Seattle, WA 98109-1024,
1 Oregon Research Institute, 1715 Franklin Boulevard, Eugene, OR 97403-1983 and
2 University of Colorado at Denver, Anthropology Department, Campus Box 103, PO Box 173364, Denver, CO 80217-3364, USA
 |
Abstract |
|---|
 Top Abstract IntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
Funding organizations increasingly want to know that successful interventions are continued after the end of a research project. Assessments of durability are rare and where done do not include the comparison communities. In this study we ascertain what tobacco control activities continued in intervention communities involved in the Community Intervention Trial for Smoking Cessation (COMMIT), a randomized, controlled community trial aimed at adult smokers, and also assessed level of tobacco control activities in the comparison communities. A mailed survey of key informants including paid staff and community volunteers in the 22 COMMIT communities was conducted. Approximately 79% of key informants responded to the survey. Although there was evidence that tobacco control activities were continuing in the intervention communities, there was an equal amount of tobacco control effort in the comparison communities. Within the specific tobacco control intervention areas, only the youth area showed more activity in intervention communities than comparison communities. We conclude that despite a positive trial outcome, differential durability was not achieved. More work needs to be done to assist communities in maintaining proven intervention activities. More study of methods to measure durability is also needed.
|
Introduction |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
Many foundations and other funding agencies expect research projects to leave something behind in the communities in which successful behavior change interventions have been conducted (Tarlov et al., 1987
; COMMIT Research Group, 1991; Altman, 1995; ASSIST Working Group on Durability, 1996). This expectation is often predicated on the assumption that successfully addressing an important community problem leads naturally into a desire by the community to continue intervention activities (Kelly, 1979; Green and McAlister, 1984; Wallack and Wallerstein, 1986; Tarlov et al., 1987; Altman et al., 1991; Jackson et al., 1994; Altman, 1995). Mobilization of a milieu, whether it be a worksite (Sorensen et al., 1990; Abrams et al., 1994; Glasgow et al., 1996), an organization (DePue et al., 1987; Elder et al., 1989; Corbett et al., 1995), a school (Scheirer, 1990; Resnicow and Botvin, 1993) or an entire town (Carlaw et al., 1983; Farquhar et al., 1985; Elder et al., 1986; COMMIT Research Group, 1991), often involves a partnership between members from the milieu and researchers. By the end of the research phase, if an intervention has been successful, partners from the community setting often wish to have the means for project durability (Altman et al., 1991; Lefebvre, 1992; Bracht et al., 1994; Lichtenstein et al., 1996). Unfortunately, many research organizations simply remove themselves, leaving it up to the community to find ways to maintain successful projects. Often, this leaves the community partners without resources, capabilities or other avenues to address the original problem. Such an action on the part of researchers is considered inconsistent with the principles of community ownership and partnership (Green and McAlister, 1984; Wallack and Wallerstein, 1986; Thompson and Kinne, 1990; Thompson et al., 1990/91).
There are many good reasons to encourage perpetuation of successful intervention results and sometimes even equivocal results. Population-level change often requires more time than is funded by an external agent; thus a small immediate change may lead to a larger change in the long run (Puska et al., 1979, 1983; Henderson et al., 1995; Beresford et al., 1997). Long-lasting and widespread behavior changes require alterations in `rules for living' (i.e. social norms) so that a new behavior is thoroughly incorporated into the society. To accomplish such a task, however, is not a trivial undertaking and often needs more time than research projects allow. The North Karelia project, for example, did not see changes in smoking behavior until 10 years after the project was initiated (Puska et al., 1979, 1983). The Community Intervention Trial for Smoking Cessation (COMMIT) did not begin to see results in smoking cessation until 3 years after the communities were randomized (COMMIT Research Group, 1995a).
To facilitate long-lasting programs, sponsoring organizations often state that their purpose in funding a particular project is to provide `seed' money until the effectiveness of a program can be assessed (Tarlov et al., 1987; COMMIT Research Group, 1991; Altman, 1995; ASSIST Working Group on Durability, 1996). In recognition that it is very costly to establish large-scale studies, sponsors typically provide initial resources which gradually dwindle as the project draws to a close. Sponsors also believe the seed money should have generated other resources by then and that a successful program should be ongoing. In an era of scarce resources, however, obtaining the resources to continue a successful program can be difficult.
Public health researchers have a responsibility to work for maintenance of successful interventions. Altman et al. (Altman et al., 1991) summarize it best when they note that research findings must be applied to have a public health effect. Part of that application is enabling communities to continue to conduct disease prevention or health promotion interventions (Jackson et al., 1994).
Only recently have a handful of researchers investigated what happens to large community projects once the external funding has been withdrawn (Bracht et al., 1994; Jackson et al., 1994; Lichtenstein et al., 1996). Few data-based studies on maintenance have been published and the existing literature focuses only on communities involved in the intervention arm of the trial and not on comparison communities that also may have made changes. For example, the Minnesota Heart Health Study reported that 60% of intervention activities were incorporated and continued in the three intervention communities 3 years after the educational intervention ended (Bracht et al., 1994). During the late 1980s and early 1990s, however, there was much emphasis on health and the Minnesota investigators pointed to a strong secular trend as the reason for reduction of risk factors in the comparison communities. Thus, it is not clear that the intervention activities maintained in the three intervention communities were similar to or different from those of the comparison communities which may have also implemented similar programs and activities. In a report on capacity-building activities for the Stanford Five-City Project, process data indicate that individuals in one community attended training sessions, and were successful in obtaining grants and becoming a model test site (Jackson et al., 1994). Again, it is not clear how the non-intervention comparison communities fared.
Previously, we reported evidence for considerable durability of tobacco control activities in the 11 intervention communities of the COMMIT trial 12 months after the end of intervention activities (Lichtenstein et al., 1996). At that time, we did not collect information on the comparison communities. In this paper, we report on the level of tobacco control activities in all 22 COMMIT communities (11 intervention and 11 comparison) 2 years after the end of the intervention phase. In this study of durability, we are particularly interested in learning whether, and to what extent, tobacco control activities in COMMIT intervention communities are of a greater magnitude than those in comparison communities where COMMIT-like intervention activities were not conducted. Comparison communities, however, were not prohibited from conducting tobacco control activities and the vast majority of communities in the US have a variety of tobacco control activities taking place regularly (Thompson and Hopp, 1991). However, we hypothesized that communities that were active participants in a comprehensive tobacco control project would have more tobacco control activities after a project ends than communities not participating in a comprehensive project.
|
Background of COMMIT |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
COMMIT was a 7-year, multi-center, randomized controlled trial designed to assess whether a comprehensive community-oriented intervention could assist smokers, especially heavy smokers, in achieving and maintaining long-term smoking cessation. Eleven pairs of communities (10 in the US and one in Ontario, Canada) participated in the trial. Each community pair was matched on demographic characteristics. After a baseline survey which determined smoking prevalence for each community and recruited evaluation cohorts of heavy, light to moderate and ex-smokers, one community in the pair was randomized to an intervention condition and one to a comparison condition (COMMIT Research Group, 1991
). The 11 intervention communities established Community Advisory Boards to guide the project activities and followed a standardized protocol which also permitted some community-specific activities (COMMIT Research Group, 1991). The National Cancer Institute (NCI) regarded the money it gave the communities as `seed money' and expected communities to contribute resources in kind to the project (COMMIT Research Group, 1991). Over 4 years of intervention, activities were targeted to health care providers, worksites, organizations, schools, media and cessation resources so that community smokers would hear regular and repeated messages about smoking cessation. The results of the trial showed that the intervention had no effect on heavy smokers but that light to moderate smokers had a statistically significant greater quit rate in the intervention communities compared to the controls (COMMIT Research Group, 1995a). Further, the quit rate was higher among less educated light and moderate smokers compared to their more educated counterparts. More detail on the trial's organization, goals and outcomes have been presented elsewhere (COMMIT Research Group, 1991, 1995a).
Durability of tobacco control activities was not originally considered an important goal of COMMIT; however, as the trial progressed, investigators and community members alike expressed interest in maintaining at least some tobacco control activities after the end of the research. The investigators and the NCI began facilitating planning for the durability of COMMIT-like activities at the beginning of the final project intervention year. By the end of the project, all 11 intervention communities had established `transition committees' to explore the possibility of continuing tobacco control activities after the research aspect of the project ended. Each transition committee wrote a plan detailing what tobacco control activities would continue and how they would be done. One year after the externally funded intervention activities ended, nine of the 11 intervention sites still had a coalition or board directing tobacco control activities, nine had dedicated funding for tobacco control and 10 had paid staff to work on tobacco control (Lichtenstein et al., 1996). Overall, there were suggestions of considerable durability of tobacco control in the intervention communities.
|
Methods |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
Defining durability
There are many ways to define durability. Some researchers have noted that durability is `institutionalization' of an intervention program into an existing organization and have developed indicators for assessing institutionalization (Goodman and Steckler, 1989; Goodman et al., 1993
). Others have measured the level to which various program parts of an intervention are retained by organizations within a community and called that `incorporation' (Bracht et al., 1994). Some researchers argue that diffusion of intervention programs to a policy level (e.g. executive orders, legislation, school policy) is necessary for durability (Altman, 1995). Yet others believe that durability requires that even more activity around a problem continue after the project ends (ASSIST Working Group on Durability, 1996). The words used to define durability vary, and include institutionalization, incorporation, capacity-building and sustainability, each with its own nuances of meaning.
In this study, we use the term `durability' to mean that some level of COMMIT-like tobacco control activities existed within the intervention communities (Thompson and Winner, 1999). Because the kinds of tobacco control activities that were promoted by COMMIT can be found in most US towns (Thompson and Hopp, 1991), we developed instruments that could be applied to both intervention and comparison communities to ascertain whether any differences existed between communities. The activities are described in more detail below.
Key informants
It was necessary to identify persons who could speak for the community about tobacco control activities. Knowledge of tobacco control activities in a community is generally found among paid employees or volunteers in specific organizations. A list of such types of organizations was developed a priori (see Table I). Informants who held occupational positions assumed to be closest to tobacco control activities (e.g. the health voluntary organizations; the health department; substance abuse programs) in the organizations were considered the `experts' about activities that were happening in the community. Key informants were identified in all 22 communities. Many of the key informants were identified through the COMMIT community analysts, individuals who maintained regular contact with individuals in intervention and comparison communities during the COMMIT trial. In cases where an individual was not known for a specific organization, telephone calls were made to establish who in the organization would be most likely to know about tobacco control activities.
|
In addition to the experts, other informants were also identified. These were individuals who had previously been involved with the COMMIT project or who were other community spokespersons. Understandably, there were many more `other' informants in the COMMIT intervention communities.
Questionnaire distribution
Mailed questionnaires were used to collect data. Reasoning that informants might know what was happening in their own area but not in others, three separate questionnaires were developed: one each for health care providers, worksite-related groups, and schools and substance abuse programs. In addition, an overall questionnaire was developed for the key informants who were expected to know about general tobacco control activities in the community. This fourth questionnaire included questions related to the following: (i) the existence and characteristics of a tobacco control coalition or group; (ii) where the latest information on tobacco control could be found; (iii) knowledge of activities of both pro- and anti-tobacco groups; (iv) tobacco control activities carried out by health care providers, worksites or health voluntary groups; (v) whether information was available for smokers on where they could get help in the community; (vi) community-wide tobacco control events that occurred (e.g. Quit & Win contests); and (vii) youth tobacco control activities.
The four intervention community questionnaires were revised for the comparison communities; specifically, questions asking about long-lasting effects of COMMIT were not included because the comparison communities had not participated in the interventions. Thus a total of eight questionnaires was developed, four for the intervention and four for the comparison communities.
After pre-testing, the cover letters and questionnaires were sent to all key informants. Three weeks later a reminder postcard was sent. If no response was received, in another 3 weeks a new packet was sent followed by a reminder postcard 3 weeks after that. Telephone calls were made to non-respondents encouraging them to return their questionnaires. If the non-responder desired, s/he could provide the information to the telephone interviewer. If key informants had left their positions, recipients of the questionnaire were asked to pass it on to the person who would be best qualified to complete it.
The questionnaire required 20–40 min to complete. The questionnaire, cover letters and the methods were approved by the Public Health Sciences Clearance Officer and the Office of Management and Budget. The study also was reviewed by Institutional Review Boards at the Fred Hutchinson Cancer Research Institute and the Oregon Research Institute. The introductory letter contained language telling recipients they were free to participate if they chose; telephone respondents gave verbal consent to participate.
Questionnaire content
The primary purpose of the questionnaire was to identify tobacco control activities that would have been similar to those promoted by the COMMIT project. We began by asking whether or not a tobacco control structure existed in the community, noting that such a structure could be a group, coalition or other entity. For those respondents who replied `yes', we ascertained the strength of the structure by combining several variables. First, we examined structures by looking at their independence or ability to set their own agenda, with independent groups defined as being stronger than those that were part of other organizations. Structures that were funded were ranked stronger than those that were not funded. Groups with paid staff were stronger than those without paid staff and structures with larger target areas of tobacco control were stronger than those with smaller areas. Responses to these variables were re-coded to reflect correct direction, summed and divided by 4 to obtain a single figure varying from 0.0 to 1.0 to indicate strength of an existing structure.
Tobacco control activities in health care settings, worksites, cessation resources, public education and with youth were examined in intervention and comparison communities.
Health care
Activities in the health care area were assessed by asking respondents to rate on a scale of 1 (not at all) to 5 (extremely) how active each of groups of hospitals, physicians, dentists, medical societies and other health professionals in their communities had been around tobacco control. The individual responses were summed and the mean calculated. In addition, respondents were asked to rate on a five-point scale (low to high) the frequency of opportunities to be trained in tobacco control for physicians, dentists, office staff of health care providers and other health professionals in the community. Respondents in the intervention communities only were asked to note on a scale of 1 (none) to 5 (a great amount) how much impact COMMIT had on tobacco control activities among health care providers in the community.
Worksites
We developed a list of activities and asked respondents to rate how often each occurred (from 1 = never to 5 = regularly) within the community in the past 2 years. The activities included passage of restrictive smoking policies in worksites, one-on-one worksite consultations for implementing smoking policies, public recognition of smoke-free workplaces and institution of stop smoking programs at workplaces. Responses to these variables were summed and divided by 4 to obtain a single mean.
Cessation resources
Respondents were asked to rate how often on a five-point scale (1 = never to 5 = regularly) community agencies offered smoking cessation sessions. Seven organizations were named including the American Cancer Society, the American Lung Association, the American Heart Association, a local hospital, Nicotine Anonymous, Seventh Day Adventist and the Health Department. Responses were summed and divided by 7 for a mean response. Informants were asked about the availability of a cessation resource guide in their community.
Public education
Respondents were asked about a number of public education activities, including local media coverage of tobacco control activities. This was measured by asking the respondent's opinion as to how active the local media had been in covering tobacco control. Level of activity was rated for newspapers, radio and television, and was ranked from never (1) to regularly (5).
Another part of public education addressed community-wide activities called `magnet events.' All respondents were asked how visible activities around the Great American SmokeOut had been in the previous year. Responses could range from 1 = not at all visible to 5 = extremely visible. Respondents were also asked if other public events around tobacco control occurred in the past 2 years.
Youth
Level of activities among youth was assessed in local schools, school-based groups, school administrators and school-based youth groups. In all cases, informants rated whether the level of activity was from very low (1) to extremely active (5). Participants were also asked the level to which the community had been active (from 1 = not at all to 5 = extremely) in enforcing policies restricting youth access to tobacco. Finally, an index was created of three activities that are commonly done to promote tobacco control among youth; these include a poster contest about the dangers of tobacco, compliance checks around sales to minors and lobbying efforts to strengthen tobacco control.
Statistical analysis
The unit of randomization and analysis was the community. A permutation test was used to compare differences between intervention and comparison communities in durability measures. For each measure, a community mean was computed and the difference between means in the matched pairs was calculated. The permutation of positive and negative signs for the 11 differences forms the permutation distribution and P values are obtained by examining the observed mean difference between the two groups against a permutation distribution (Edgington, 1987). In three measures, there were two communities without data and means were imputed by using the mean value of the community means within the particular treatment arm. In this way, communities did not have to be dropped from the analysis.
|
Results |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
Responses were received from 329 of the 418 informants surveyed for an overall response rate of 79%. Because former COMMIT staff were some of the informants in intervention communities, the median number of people interviewed was higher for intervention (median N = 18; range = 16–24) than comparison (median N = 11; range = 4–19) communities. Intervention communities had a slightly higher response rate (80%; range = 67–95%) than comparison communities (75%; range = 50–90%). The number of respondents in the Expert Set was 97 in combined intervention communities and 120 in combined comparison communities.
Respondents were asked whether or not a tobacco control structure existed in their community; if respondents said `yes,' they were asked to assess the strength of that structure. In 18 of the communities (nine intervention and nine comparison), there was agreement by more than 50% of the Expert Set that a tobacco control structure existed. As shown in Table II, there was little difference between intervention and comparison communities in the overall percentage of respondents who agreed that a structure existed, whether examining the responses given by all individuals surveyed or by the Expert Set respondents. In the intervention communities, there was good agreement between the entire set of respondents and the Expert Set, except for pairs `F' and `H', where the entire set of respondents were more likely than the Expert Set respondents to state that a structure existed. In the comparison communities, general and Expert Set respondents exhibited more disagreement than the intervention group. In pair `A', less than half of all respondents thought that a structure existed, while all of the set experts were positive in their responses. Pairs `F' and `I' also have substantial discordance. When examining the two groups overall, the range of percentages around the grouped mean is similar for all respondents and Expert Set respondents in the intervention communities (range = 31.5–95.2 and 0.0–100.0, respectively) and the comparison communities (range = 42.9–100.0 and 0.0–100.0, respectively).
|
In both arms of the trial, the evaluation of the strength of the structure is relatively high (see Table II
). Both arms of the trial showed good agreement among all respondents and the Expert Set respondents regarding the strength of the structure, with all respondents in the intervention communities indicating higher strength (0.61) than the Expert Set respondents (0.56). The comparison communities had a slightly lower overall strength reported by all respondents (0.54) than by the Expert Set (0.58).
Current activity levels in the areas targeted by COMMIT (e.g. health care, worksites, public education) are shown in Table III. The level of activity is similar across the community pairs. Intervention communities show slight gains in the areas of `availability of smoking cessation information through a cessation resources guide' and `combined youth events'. `Presence of other community cessation events' and `enforcement of prohibitions against youth smoking' were significantly higher in the intervention communities. On the other hand, comparison communities have small advantages in `smoking cessation activities' and `visibility of the Great American SmokeOut'.
|
Durability also was assessed by examining knowledge about availability of current tobacco control information and the presence of active anti-tobacco groups in the community. Both intervention and comparison communities gave similar responses to a question asking where the respondent would go to obtain current tobacco control information (see Table IV
). However, the comparison communities placed more emphasis on receiving information from the three health care voluntary agencies: American Cancer Society, the American Lung Association and the American Heart Association. In addition, comparison communities were more likely to obtain information from their health departments. Other anti-tobacco groups in all of the communities were sparse. Doctors Ought to Care had double the visibility in intervention communities, while Action on Smoking and Health and Americans for Non-smokers' Rights were more visible in the comparison communities.
|
Finally, intervention community respondents only were asked to assess the overall community impact remaining from COMMIT. The overall mean response was 3.4 on a scale of 1 to 5 where 1 meant `no impact' and 5 meant `a great amount'. That question was not asked of respondents in the comparison communities.
|
Discussion |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
The COMMIT trial showed a significant increase in smoking cessation rates among light-to-moderate smokers in intervention communities. The majority of the COMMIT intervention communities displayed strong interest in continuing tobacco control activities. To estimate the breadth and intensity of tobacco control efforts after COMMIT, key informants within the 22 communities that participated in the COMMIT project were surveyed 2 years after the project intervention phase ended. This produced a snapshot of ongoing tobacco control activities both in the intervention and the comparison towns.
An earlier study conducted only within the COMMIT intervention communities found that a great deal of the tobacco control activity instigated by COMMIT was continuing (Lichtenstein et al., 1996). That study focused on tobacco control 12–16 months post-intervention and before the COMMIT outcome data were reported. Despite not knowing whether the trial had been successful, nine of the 11 intervention communities had coalitions or boards in place, nine had dedicated funding and 10 had some paid staff time. There was also evidence of activities focused on youth, who were only a minor target in the original project because of the emphasis on heavy smokers. The findings of that study supported other findings of large-scale community studies that suggest that durability can be attained (Bracht et al., 1994; Jackson et al., 1994).
Data from this study substantiated the earlier findings in the intervention communities. Two years after the intervention ended, nine of the 11 intervention communities had coalitions or boards. Respondents from six of the communities scored strength of the coalition (resources, paid staff, and reach of activities) as 0.75 or higher on a scale that ranged from 0.0 to 100.0. Taken by itself, this might be considered similar to the durability identified in intervention communities in the Minnesota Heart Health Program (Bracht et al., 1994) and the Stanford Five-City Study (Jackson et al., 1994). We had the opportunity to evaluate the data relative to data from the comparison communities. There are few differences in scores between the intervention and comparison communities. The differences that do appear are small. Only two activities, `presence of other community cessation events' and `enforcement of prohibitions against youth smoking', showed significant differences between intervention and comparison communities.
There are a number of potential explanations for this overall outcome. As with other large-scale community trials, COMMIT saw a large secular trend that affected tobacco control activities throughout the country. California had passed its proposition increasing tobacco taxes and dedicating those taxes to tobacco control activities. Massachusetts was ready to do the same. Both of those states had COMMIT sites.
Another explanation might lie in the American Stop Smoking Intervention Study for Cancer Prevention (ASSIST). ASSIST is a collaboration between state health departments and the American Cancer Society designed to build coalitions in 17 states that will conduct tobacco control activities in specified channels of intervention to reduce smoking prevalence. The ASSIST project began before the COMMIT project was completed and some of the COMMIT communities worked intensely to become an ASSIST site. That may have affected results of this durability study in two ways. Comparison communities in ASSIST states might have received tobacco control funds and so were able to develop their own tobacco control infrastructure. On the other hand, COMMIT communities not awarded ASSIST funds may have struggled with maintaining even a low level of tobacco control when their communities were not awarded ASSIST funds.
Another explanation may be that COMMIT did not spend enough time planning for durability. The topic of durability was raised only in the last 18 months of the trial. At that time, many of the community Boards were concerned with the ending of COMMIT and wanted help in planning for the future. The communities received little or no training or technical assistance in generating plans for the future or in finding funds to sustain some activities. Some communities did hold workshops or retreats to plan for the future, but appeared to be overwhelmed by the difficulty of finding additional funding to establish a long-term infrastructure. Further, although a goal of COMMIT was to increase the community capacity to conduct tobacco control activities, the project emphasized that it was not taking over activities from existing tobacco control organizations; rather, its goal was to increase the abilities of those organizations to work together with other community groups in an effective and coordinated manner. This may have caused confusion as to long-term goals and planning for life after COMMIT. It was not until midway through the intervention that researchers and community members agreed that durability might be an important feature to include in COMMIT. This did not leave sufficient time to engage in trial-wide planning to keep tobacco control activities going.
Typically the organizations that deal with cancer or smoking-related diseases in communities are the health voluntary groups (American Cancer Society, American Lung Association and American Heart Association) and, to some extent, health departments, and the assumption was that these organizations would build capacity to provide smoking-related services and information. Interestingly, when asked where information could be obtained, the major differences between intervention and comparison communities was less use of the health voluntary groups in the intervention communities. In retrospect, this is not so surprising. Health voluntary agencies have limited resources. COMMIT had the funds to advertise its presence specifically for tobacco control; thus, it acquired the reputation for being the place to go for tobacco-related information. In the meantime, the health voluntary agencies saw COMMIT as being able to free the organizations to pursue other cancer-related activities. When the COMMIT project ended, the voluntary health groups may not have taken up the activities again.
In addition to these difficulties in planning for durability, staff and volunteers had to turn their attention from intervention activities to fund raising for any durability to occur. Fund-raising is an arduous task at best and the COMMIT staff had numerous project-related close-out activities to conduct. This may have led to reduced energy to deal with durability of tobacco control activities.
The COMMIT project required that funds given to the communities be spent annually on tobacco control activities. It was not possible, under the conditions of the contract, to save money for continuing tobacco control activities after the trial was over. Some communities were able to obtain donated space from a hospital, voluntary health organization, health department or other source; however, the majority of the communities had few resources remaining from COMMIT. Given this constraint, it may be understandable that few differences existed between intervention and comparison communities 2 years after the trial.
This study focused on a quantitative assessment of activities; therefore, it is not clear that the quality of tobacco control activities was the same in intervention and comparison communities. A more qualitative assessment might have identified differences in activities between the communities. For example, the cessation resource guides in the intervention communities were comprehensive, colorful and had numerous self-help instructions on quitting, compared to a list of where one could go to get help on quitting commonly seen in the comparison communities. Qualitatively, this could have a different impact on smokers. Future research may wish to include qualitative examinations of durability.
The two significant differences in durability activities were `presence of other community cessation events' and `enforcement of prohibitions against youth smoking'. The other community activities were overwhelmingly `Quit & Win' contests. Those contests had been an active part of COMMIT and many intervention communities found ways to continue them by linking up with other groups in the community.
The emphasis on youth prevention in the intervention communities increased substantially after the intervention ended. From the beginning of the trial, researchers and community members were encouraged to target heavy smokers and since few youth fell into that category, little emphasis was placed on them initially. As the trial planning progressed, activities directed at youth or with youth were added to the protocol. The protocol of youth activities, however, was later downsized because of budget consideration. By the end of the trial, many communities, recognizing that prevention now could decrease the need for cessation later, wanted to focus on youth. The data likely reflect this interest as the intervention communities exceed the comparison communities in all three areas of youth activities, although only enforcement is significant.
This study on durability, although breaking new ground by examining comparison as well as intervention communities, has a number of limitations. The sample for this survey was selected on the basis of occupational positions in each of the 22 communities. It is possible that by using the positional approach, some other key individuals may have been missed. Respondents in our intervention communities may have been more knowledgeable about activities that were going on in the community because their level of awareness of tobacco control had been raised by COMMIT. This may have been particularly true in the area of youth where the issue of prevention became clear as the trial went on and where many intervention communities focused their energies once the trial was over. Being more aware about tobacco control could have led to reporting more tobacco control activities than the comparison communities. Alternatively, tobacco control activities could have been under-reported as respondents in intervention communities compared current activities with the past COMMIT years where there had been a much higher rate of activity.
|
Conclusion |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
This study contributes to the growing body of literature on the durability of intervention activities instigated during community trials. It is the first study to compare ongoing activities in both intervention and comparison communities after a research study has ended. We found considerable evidence for durability in the intervention communities, with the overwhelming majority of them having coalitions or advisory groups to conduct tobacco control activities. There were, however, only small differences between intervention and comparison communities in the level of tobacco control activities after the funding period ended.
A number of lessons can be gleaned from this study. First, the time to establish a solid infrastructure for durability probably requires more than the last few months or the last year of a project. Durability planning should begin when project activities are implemented so that their maintenance will be natural when the external funds decrease. Secondly, communities in general want their project activities to continue and should be encouraged to do so when a trial has a successful outcome. Finally, to translate research data to the real world, we must understand durability and the factors related to durability so that other communities can benefit from research projects.
|
Acknowledgments |
|---|
This study was supported by contract no. CN64100 from the National Cancer Institute.
|
References |
|---|
|
TopAbstractIntroductionBackground of COMMITMethodsResultsDiscussionConclusionReferences |
|---|
Abrams, D. B., Boutwell, B. W., Grizzle, J., Heimendinger, J., Sorensen, G. and Varnes, J. (1994) Cancer control at the workplace: the Working Well Trial. Preventive Medicine, 23, 15–27.[Web of Science][Medline]
Altman, D. G. (1995) Sustaining interventions in community systems: on the relationship between researchers and communities. Health Psychology, 14, 526–536.[Web of Science][Medline]
Altman, D. G., Endres, J., Linzer, J., Lorig, K., Howard-Pitney, B. and Rogers, T. (1991) Obstacles to and future goals of ten comprehensive community health promotion projects. Journal of Community Health, 16, 299–314.[Medline]
ASSIST Working Group on Durability (1996) Turning Point for Tobacco Control: Toward a National Strategy to Prevent and Control Tobacco Use. Prospect Associates, Rockville, MD.
Beresford, S. A., Curry, S. J., Kristal, A. R., Lazovich, D., Feng, Z. and Wagner, E. H. (1997) A dietary intervention in primary care practice: the Eating Patterns Study. American Journal of Public Health, 87, 610–616.
Bracht, N., Finnegan, J. R., Rissel, C., Weisbrod, R., Gleason, J., Corbett, J. and Veblen-Mortenson, S. (1994) Community ownership and program continuation following a health demonstration project. Health Education Research, 9, 243–255.
Carlaw, R. W., Mittelmark, M. B., Bracht, N. and Luepker, R. (1984) Organization for a community cardiovascular health program: experiences from the Minnesota Heart Health Program. Health Education Quarterly, 11, 243–252.[Web of Science][Medline]
Carleton, R. A., Lasater, T. M., Assaf, A. R., Feldman, H. A., McKinlay, S. and the Pawtucket Heart Health Program Writing Group (1995) The Pawtucket Heart health Program: community changes in cardiovascular risk factors and projected disease risk. American Journal of Public Health, 85, 777–785.
COMMIT Research Group (1991) Community Intervention Trial for Smoking Cessation (COMMIT): summary of design and intervention. Journal of the National Cancer Institute, 83, 1620–1628.
COMMIT Research Group (1995a) Community Intervention Trial for Smoking Cessation (COMMIT): I. Cohort results from a four-year community intervention. American Journal of Public Health, 85, 183–192.
COMMIT Research Group (1995b) Community Intervention Trial for Smoking Cessation (COMMIT): II. Changes in adult cigarette smoking prevalence. American Journal of Public Health, 85, 193–200.
Corbett, K. K., Nettekoven, L., Churchill, L. C., Dalton, L. T., Johnson, C. L., Dickinson, L., Sorensen, G. and Thompson, B. (1995) Involving diverse community organizations in tobacco control activities. In Community-Based Interventions for Smokers: The COMMIT Experience. Smoking and Tobacco Control. Monograph no. 7. USDHHS, Public Health Service, National Institutes of Health, National Cancer Society, Bethesda, MD.
DePue, J. D., Wells, B. J., Lasater, T. M. and Carleton, R. A. (1987) Training volunteers to conduct heart health programs in churches. American Journal of Preventive Medicine, 3, 51–57.[Web of Science][Medline]
Edgington, E. S. (1987) Randomization Tests, 2nd edn. Marcel Dekker, New York.
Elder, J. P., McGraw, S. A., Abrams, D. B., Ferreira, A., Lasater, T. M., Longpre, H., Peterson, G. S., Schwertfeger, R. and Carleton, R. A. (1986) Organizational and community approaches to community-wide prevention of heart disease: the first 2 years of the Pawtucket Heart Health Program. Preventive Medicine, 15, 107–117.[Web of Science][Medline]
Elder, J. P., Sallis, J. F., Mayer, J. A., Hammond, N. and Perelinski, S. (1989) Community-based health promotion: a survey of churches, labor unions, supermarkets, and restaurants. Journal of Community Health, 14, 159–168.[Medline]
Farquhar, J. W., Fortmann, S. P., Maccoby, N., Haskell, W. L., Williams, P. T., Flora, J. A., Taylor, C. B., Brown, B. W., Solomon, D. S. and Hulley, S. B. (1985) The Stanford Five-City Project: design and methods. American Journal of Epidemiology, 122, 323–334.
Glasgow, R. E., Sorensen, G., Giffen, C., Shipley, R. H., Corbett, K. K., Lynn, W. for the COMMIT Research Group. (1996) Promoting worksite smoking control policies and actions: the COMMIT experience. Preventive Medicine, 25, 186–194.[Web of Science][Medline]
Goodman, R. M. and Steckler, A. M. (1987) A model for the institutionalization of health promotion programs. Family and Community Health, 11, 63–78.
Goodman, R. M., McLeroy, K. R., Steckler, A. M. and Hoyle, R. (1993) Development of level of institutionalization scales for health promotion programs. Health Education Quarterly, 20, 161–178.[Web of Science][Medline]
Green, L. W. and McAllister, A. (1984) Macro-intervention to support health behavior change: some theoretical perspectives and practical reflections. Health Education Quarterly, 11, 322–339.[Medline]
Henderson, M., Thompson, B. and Kristal, A. (1995) Behavioural intervention versus chemoprevention. In Hakama, M., Beral, V., Buiatti, E., Faivre, J. and Parkin, D. M. (eds), Chemoprevention in Cancer Control (IARC Scientific Publication no. 136). IARC, Lyon, France, pp. 123–130.
Jackson, C., Fortmann, S. P., Flora, J. A., Melton, R. J., Snider, J. P. and Littlefield, D. (1994) The capacity-building approach to intervention maintenance implemented by the Stanford Five-City Project. Health Education Research, 9, 385–396.
Kelly, J. G. (1979) T'ain't what you do, it's the way you do it. American Journal of Community Psychology, 7, 239–261.[Web of Science][Medline]
Lando, H. A., Pechacek, T. F., Pirie, P. L., Murray, D. M., Mittelmark, M. B., Lichtenstein, E., Nothwehr, F. and Gray, C. (1995) Changes in adult cigarette smoking in the Minnesota Heart Health Program. American Journal of Public Health, 85, 201–208.
Lefebvre, R. C. (1992) Sustainability of health promotion programmes. Health Promotion International, 7, 239–240.
Lichtenstein, E., Thompson, B., Nettekoven, L., Corbett, K. for the COMMIT Research Group. (1996) Durability of tobacco control activities in eleven North American communities: life after the Community Intervention Trial for Smoking Cessation (COMMIT). Health Education Research, 11, 527–534.
Puska, P., Tuomilehto, J., Salonen, J., Neittaanmaki, L., Maki, J., Virtamo, J., Nissinen, A., Koskel, K. and Takalo, T. (1979) Changes in coronary risk factors during comprehensive five-year programme to control cardiovascular disease (North Karelia Project). British Medical Journal, ii, 1173–1178.
Puska, P., Salonen, J. T., Nissinen, A. and Tuomilehto, J. (1983) Ten years of the North Karelia Project: results with community-based prevention of coronary heart disease. Scandinavian Journal of Social Medicine, 11, 65–68.[Web of Science][Medline]
Resnicow, K. and Botvin, G. (1993) School-based substance use prevention programs: why do effects decay? Preventive Medicine, 22, 484–490.[Web of Science][Medline]
Scheirer, M. (1990) The life cycle of an innovation: adoption versus discontinuation of the fluoride mouth rinse program in schools. Journal of Health and Social Behavior, 31, 203–215.[Web of Science][Medline]
Sorensen, G., Glasgow, R. E. and Corbett, K. K. (1990) Involving worksites and other organizations. In Bracht, N. (ed.), Health Promotion at the Community Level. Sage, Newbury Park, CA, pp. 158–184.
Tarlov, A. R., Kehrer, B. H., Hall, D. P., Samuels, S. E., Brown, G. S., Felix, M. R. and Ross, J. A. (1987) Foundation work: the health promotion program of the Henry J. Kaiser Family Foundation. American Journal of Health promotion, 2, 74–80.[Medline]
Thompson, B. and Hopp, H. P. (1991) Community-based programs for smoking cessation. Clinics in Chest Medicine, 12, 801–818.[Web of Science][Medline]
Thompson, B. and Kinne, S. (1990) Theories of community change: review, synthesis, and application. In Bracht, N. (ed.), Health Promotion at the Community Level. Sage, Newbury Park, CA, pp. 45–65.
Thompson, B. and Winner, C. (1999) Durability of community intervention programs: definitions, empirical studies, and strategic planning. In Bracht, N. (ed.), Health Promotion at The Community Level: New Advances, 2nd edn. Sage, Thousand Oaks, CA, pp. 137–154.
Thompson, B., Lichtenstein, E., Wallack, L. and Pechacek, T. (1990/91) Principles of community organization and partnership for smoking cessation in the Community Intervention Trial for Smoking Cessation (COMMIT). International Quarterly of Community Health Education, 11, 187–203.
Wallack, L. and Wallerstein, N. (1986) Health education and prevention: design community initiatives. International Quarterly of Community Health Education, 7, 319–342.
Received on August 3, 1998; accepted on May 17, 1999
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. A. Scheirer Is Sustainability Possible? A Review and Commentary on Empirical Studies of Program Sustainability American Journal of Evaluation, September 1, 2005; 26(3): 320 - 347. [Abstract] [PDF]
|
|
|
|
 |
|
 B. L. Riley, S. M. Taylor, and S. J. Elliott Organizational capacity and implementation change: a comparative case study of heart health promotion in Ontario public health agencies Health Educ. Res., December 1, 2003; 18(6): 754 - 769. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. M. Reininger, M. Vincent, S. F. Griffin, R. F. Valois, D. Taylor, D. Parra-Medina, A. Evans, and M. Rousseau Evaluation of Statewide Teen Pregnancy Prevention Initiatives: Challenges, Methods, and Lessons Learned Health Promot Pract, July 1, 2003; 4(3): 323 - 335. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||