Vous êtes sur la page 1sur 5

& Tobacco Research Advance Access published July 7, 2011 Nicotine & Nicotine Tobacco Research

Brief report

Behavior Change Techniques Used in Group-Based Behavioral Support by the English Stop-Smoking Services and Preliminary Assessment of Association with Short-term Quit Outcomes
Robert West, B.Sc., Ph.D.,1 Adam Evans, B.Sc.,2 & Susan Michie, B.A., M.Phil., D.Phil.3
Cancer Research UK Health Behaviour Research Centre, Department of Epidemiology and Public Health, University College London and the NHS Centre for Smoking Cessation and Training, London, UK 2 Department of Psychology, University of Surrey, Guildford, UK 3 UCL Division of Psychology and Language Sciences, University College London and the NHS Centre for Smoking Cessation and Training, London, UK
1
Downloaded from ntr.oxfordjournals.org at Society for Research on Nicotine and Tobacco member access on August 8, 2011

Corresponding Author: Robert West, Cancer Research UK Health Behaviour Research Centre, Department of Epidemiology and Public Health, University College London and the NHS Centre for Smoking Cessation and Training, 216 Torrington Place, London WC1E 6BT, London, UK. E-mail: robertwest100@gmail.com Received November 20, 2010; accepted May 9, 2011

Abstract
Objective: To develop a reliable coding scheme for com ponents of group-based behavioral support for smoking cessation, to establish the frequency of inclusion in English Stop-Smoking Service (SSS) treatment manuals of specific components, and to investigate the associations between inclusion of behavior change techniques (BCTs) and service success rates. Methods: A taxonomy of BCTs specific to group-based behavioral support was developed and reliability of use assessed. All English SSSs (n = 145) were contacted to request their groupsupport treatment manuals. BCTs included in the manuals were identified using this taxonomy. Associations between inclusion of specific BCTs and short-term (4-week) self-reported quit outcomes were assessed. Results: Fourteen group-support BCTs were identified with >90% agreement between coders. One hundred and seven services responded to the request for group-support manuals of which 30 had suitable documents. On average, 7 BCTs were included in each manual. Two were positively associated with 4-week quit rates: communicate group member identities and a betting game (a financial deposit that is lost if a stop-smoking buddy relapses). Conclusion: It is possible to reliably code group-specific BCTs for smoking cessation. Fourteen such techniques are present in guideline documents of which 2 appear to be associated with higher short-term self-reported quit rates when included in treatment manuals of English SSSs.

Introduction
Treatment to aid cessation, in the form of a combination of behavioral support and medication, is effective and highly costeffective as a life-saving clinical intervention (Parrott, Godfrey, Raw, West, & McNeill, 1998; U.S. Department of Health and Human Services, 2008; West, McNeill, & Raw, 2000). A national network of Stop-Smoking Services (SSSs) providing this treatment was introduced in England in 1999 (Department of Health, 1999). The National Health Service (NHS) is currently organized around 145 primary care trusts (PCTs; though the number changes frequently with merging and splitting), each of which has autonomy to fund and configure its health services under general guidance from national bodies, such as the National Institute for Clinical and Healthcare Excellence and the Department of Health. This autonomy can lead to wide variation in practice. The SSSs are funded and organized by these PCTs. Although use of NHS SSSs generally increases smokers chances of quitting (Ferguson, Bauld, Chesterman, & Judge, 2005; Judge, Bauld, Chesterman, & Ferguson, 2005), success rates vary considerably. This will be due in part to variation in smoker characteristics and reporting practices (Bauld, Chesterman, Judge, Pound, & Coleman, 2003) but may also result from variation in delivery of the services. We recently reported that inclusion of specific behavior change techniques (BCTs) in treatment manuals for individual behavioral support could account for a significant proportion of the variance (West, Walia, Hyder, Shahab, & Michie, 2010). A significant proportion of treatment in the United Kingdom and globally is provided in groups, so it is important to know what BCTs are used in groupbased behavioral support and whether any associations can be

doi: 10.1093/ntr/ntr120 The Author 2011. Published by Oxford University Press on behalf of the Society for Research on Nicotine and Tobacco. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Group-based BCTs and smoking cessation


found between inclusion of specific BCTs in treatment manuals and outcomes. This can inform the development of manuals and identification of competences for stop-smoking practitioners. It also provides a basis for further research to improve the effectiveness of behavioral support. Treatment manuals for smoking cessation support specify the number and structure of sessions and how each session is to be run and what topics should be covered, and activities undertaken, by the practitioner following national recommendations (see http://www.ncsct.co.uk). All licensed medications to aid cessation are available as part of the treatment program usually for between 8 and 12 weeks. Sessions usually start a week before the quit date and are run weekly so that six sessions have been attended by the 4th week after the quit date. As with individual behavioral support, there may be a significant gap between what is specified in manuals and what may occur in practice, but to the extent that the manuals act as a guide to the support provided, it may be possible to detect an association between inclusion of specific BCTs and success rates in the face of this noise. The BCTs were identified in this study in a similar manner to that used in the individual behavioral support study: They were arrived at inductively. They were then grouped by putative function using PRIME Theory, a theory of motivation that has been applied to addiction and more specifically to smoking (West, 2009). This theory identifies several goals for behavioral support: (a) addressing motivation by minimizing the frequency and strength of momentary desire to engage in the old behavior and maximizing the countervailing desire to sustain the new behavior pattern and (b) maximizing the selfregulatory capacity and skills needed to achieve this, for example, by advising on ways to avoid smoking cues (West, 2009). BCTs may act directly on these mechanisms or act indirectly by (c) promoting adjuvant activities, for example, taking smoking cessation medication and (d) supporting the other BCTs, for example, by establishing rapport. It has been proposed that group-based behavioral support can address the motivational goals by fostering team building, which creates a desire to win team approval and not let down others in the group (Hajek, 1989). On this basis, it was hypothesized that the BCTs that foster group cohesion would contribute significantly to success rates. This study formed part of a program of research carried out by the National Health Service Centre for Smoking Cessation and Training (NCSCT). The NCSCT aims to discover bestpractice supporting smoking cessation, identify the competences required of stop-smoking practitioners, and develop and implement assessment and training to ensure that all specialists possess those competences (see http://www.ncsct.co.uk). the previously identified individual BCTs. Each sentence in the guidance documents was discussed by two researchers and, where appropriate, given a BCT label. BCTs, which were judged to have the same specific function, were then grouped into categories. For example, a ceremonial throwing away and a celebration toast could be considered as group bonding tasks. Detailed and specific definitions were formulated, and a draft coding manual was developed. The procedure and results were reviewed by three experts in smoking cessation (RW, Peter Hajek, and Andy McEwen) and a taxonomy expert (SM), and amendments were made in response to their feedback. Having developed a draft taxonomy, the next step was to assess the interrater reliability (IRR) of applying this to specify the intervention content of service manuals. All 145 SSSs in England at the time of the study were requested on up to six occasions to send their treatment manuals for group-based behavioral support. The request was for any documents that provided guidance or instructions on the content of the interactions with smokers. Manuals were assessed as to whether they contained sufficient detail to identify BCTs; this was done by two researchers independently with 100% agreement. IRR was assessed on two sets of three randomly selected service manuals. If one coder identified a BCT in a section of text, the coding from the other coder was examined. If this coder identified the same BCT, the agreement was registered. If no BCT was identified or a different BCT was identified, disagreement was registered. Additionally, for each manual, the number of times each BCT was mentioned was recorded by both raters. There was discussion where disagreement occurred, and wording of the taxonomy was changed to make definitions clearer. The use of the refined taxonomy on the second set of manuals resulted in a satisfactory IRR and no uncertainties over definitions, so no further changes were made. The remaining manuals were coded using this taxonomy. The instructions for coding and the BCT labels and definitions in the coding guidelines are provided in the supplementary material. The number of times that each of the 14 BCTs was mentioned in each manual was recorded. Outcome data for each SSS were obtained from the English Department of Health for the year in which the manuals were provided (20082009). Each SSS is required to provide to the Department of Health the number of smokers setting a quit date with the service and the 4-week quit outcomes by smokers self-report. These outcomes consisted of a claim by the client not to have smoked at all for the 2 weeks prior to this point: That is, there is a 2-week grace period immediately after the quit date during which slips are allowed. Short-term outcome data of this kind do not translate directly into long-term abstinence, but there are now sufficient data to be able to make robust projections (Ferguson et al., 2005; Hughes, Keely, & Naud, 2004). On average, smokers who remain abstinent at 4 weeks have a 30% chance of remaining abstinent for 12 months, and 12-month abstainers are estimated to have a 70% chance of remaining abstinent permanently (Etter & Stapleton, 2006). Ideally, we would have used quit rates verified by expiredair carbon monoxide (CO) concentrations as we had done for the study of individual behavioral support. However, these data were not available for those smokers receiving group-based support.

Downloaded from ntr.oxfordjournals.org at Society for Research on Nicotine and Tobacco member access on August 8, 2011

Methods
Two smoking cessation guidance documents (Hajek, 2010; McEwen, Hajek, McRobbie, & West, 2006) were analyzed by two coders into group-specific BCTs. The approach adopted was similar to that previously used for individual support (Abraham & Michie, 2008; West et al., 2010). A group-specific BCT was defined as any explicit description of intervention content that can alter a participants smoking behavior, which can only be conducted in a group context. These occur alongside

Nicotine & Tobacco Research


Contacted (n= 145, 100%)

Responded (n= 107, 73.8%)

Didnt Respond (n= 38, 26.2%)

was the PCT, and the dependent variable was success rate of that PCT. The set of BCTs found to be associated with higher success rates was then entered into a forward stepwise multiple regression to determine any independent contribution. Using only one datapoint per SSS is a conservative analysis, but data were not available to enable us to undertake the analysis using individual outcome data from clients.

Run group sessions (n= 67, 46.2%)

Do not run group sessions (n= 40, 27.6%)

Results
Agreement between the two coders was 95.3% for the first pilot sample, 92.4% for the second pilot sample, and 94.2% for the combined pilot samples. Cohens kappa coefficients were .74 for both pilot samples and when combined (p < .001 in all cases). The final taxonomy of group-specific BCTs contained 14 categories. The mean number of BCTs identified in each manual was 7.16 (range: 112; SD = 3.45).
Downloaded from ntr.oxfordjournals.org at Society for Research on Nicotine and Tobacco member access on August 8, 2011

Written manual (n= 39, 26.9%)

No written manual (n= 28, 19.3%)

Manuals suitable for coding (n= 31, 21.4%)

Used 1-2-1 session manual (n= 4, 2.8%)

Manuals without sufficient information to identify techniques (n= 4, 2.8%)

Manuals suitable for coding with quit rate data available (n= 30, 20.7%)

Figure 1. Flowchart of the response rate and inclusion criteria for analysis. Percentages are of the total number of primary care trusts contacted.

Percentage agreement between coders and a weighted kappa were calculated to assess IRR. The proportion of SSS manuals that incorporated each BCT was calculated. The success rates of SSSs (i.e., the percentage of smokers who were 4-week successes) that included or did not include each BCT were compared by analysis of variance. Note that the unit of analysis

Of the 107 (73.8%) SSSs who responded to the request for treatment manuals, 67 (62.6% of the total) ran group sessions (from government statistics, the total nationally at the time was 104), but only 30 (20.7% of the total) had a manual suitable for coding and had outcome data available (Figure 1). The mean number of sessions identified in the manuals was 6.6 (range: 58; SD = 0.65). Table 1 shows for each BCT the percentage of SSSs that included the BCT in its manual. Encourage group discussion (93.3%), Communicate group member identities (80%), and

Table 1.The Number of Manuals Each BCT Appeared in and the Mean (SD) Percentage Success Rate Where They Were Present or Absent
BCT 1. Screen for suitability for group support 2. Explain group support 3. Communicate group member identities 4. Use furniture to reinforce group interaction 5. Encourage comparison of cabon monoxide readings 6. Report on missing members 7. Emphasize each individual has a responsibility for the group 8. Encourage mutual support 9. Group tasks that promote interaction and/or bonding 10. Encourage group discussion 11. Implement buddy system 12. Betting game (placing a financial deposit which is lost if a stop smoking buddy relapses) 13. Public promise 14. Discuss maintenance support Number of manuals Mean (SD) percent 4-week Mean (SD) percent 4-week p Value for difference in BCT appeared in (%) quitters when BCT present quitters when BCT absent mean percent success 14 (46.7) 22 (73.3) 24 (80.0) 5 (16.7) 0 (0) 9 (30.0) 11 (36.7) 23 (76.7) 20 (66.7) 28 (93.3) 16 (53.3) 7 (23.3) 60 (10.9) 62 (8.3) 62 (9.8) 58 (9.4) a 64 (10.4) 61 (11.1) 59 (10.5) 63 (10.2) 64 (10.2) 67 (6.1) 61 (10.7) 57 (15.4) 53 (10.7) 61 (10.9) 59 (10.6) 60 (10.6) 64 (11.2) 55 (9.8) 57 (10.1) 58 (10.8) .81 .28 .04 .61 .27 .77 .38 .05

.07 .04

18 (60.0) 18 (60.0)

63 (8.3) 59 (11.1)

57 (13.0) 63 (9.5)

.15 .24

Note. BCT = behavior change technique. a No comparison made when fewer than 10% or more than 90% of services used the BCT.

Group-based BCTs and smoking cessation


Encourage mutual support (76.7%) were the most commonly included BCTs. Encourage comparison of CO readings (0%), Use furniture to reinforce interaction (16.7%), and Betting game (23.3%) were the least used BCTs. No additional BCTs were identified. Only 12 of the 14 BCTs showed sufficient variation across SSSs to examine associations with outcome. Two of these were positively and significantly associated with 4-week self-report quit outcomes (communicate group member identities and betting game), and one was marginally significant (group tasks that promote interaction/bonding; Table 1). On entering these three variables into a forward stepwise multiple regression, only the first two made independent contributions to predicting success rates (B = 0.09, p = .042 and B = 0.08, p = .049, respectively). There was no association between the number of BCTs in treatment manuals and success rates. work is a starting point and basis for including other BCTs as and when these are identified in other countries or in group treatment progresses. Finally, we were not able to allow for other possible confounding variables in the analysis of associations with quit outcome. For example, we did not record individualfocused BCTs that would also have been present (e.g., advising on use of medication). Neither is it known what proportion of those who set a quit date never made it to that date (these are counted as treatment failures). Neither did we have access to other individual-level data on the smokers. Limitations of the data mean that the findings must be considered preliminary, but establishing this proof of principle is vital to starting a systematic analysis of this important area. Future research should examine whether the number of BCT categories needs to be extended and refined, the components of group-based support programs evaluated in randomized controlled trials, and associations between BCTs specified in treatment manuals and those delivered in practice and between those delivered in practice and treatment outcomes.

Downloaded from ntr.oxfordjournals.org at Society for Research on Nicotine and Tobacco member access on August 8, 2011

Discussion
The results showed that it was possible to devise a taxonomy that could be used reliably to classify group-specific BCTs contained in manuals for behavioral support to assist with smoking cessation. Fourteen such BCTs were identified. Of these, communicating group member identities and a betting game could be shown to be associated with higher success rates. While the former featured in 80% of treatment manuals, the latter featured in only 23%. This is first time, to our knowledge, that an attempt has been made to systematically identify the components of groupbased behavioral support for smoking cessation. The fact that it can be done reliably potentially provides a basis for examining variation in practice, establishing training programs, and assessing associations with outcomes. The study had a number of limitations, and the findings must be considered preliminary. First, the fact that only 30 of 104 services that provided group treatment provided usable treatment manuals raises the possibility that these were unrepresentative. However, the self-reported 4-week success rates in these 30 were very similar to the national average (60% vs. 61%). Second, it is possible that some services would have continued group support beyond 4 weeks, which might affect longer term outcomes. However, to use data beyond that point would have led to confounding of treatment duration. The treatment process is such that after the first week of abstinence, the group sessions follow the same format so that introduction of some of the BCTs after the 4th week would not have occurred. Third, the findings relating to success rates relied on self-report data, which were recorded by the treatment provider, and we cannot rule out that BCTs may have raised the deception rate rather than the true quit rate. Fourth, the power to detect associations between BCTs and outcome was low with a sample size of only 30 services. A 5th limitation is that we initially identified BCTs using two manuals that formed the basis for treatment recommendations in the United Kingdom; we are not aware of any other such manuals in the United Kingdom. We also added further BCTs that emerged during the examination of individual service treatment manuals. However, there may be other BCTs in use in other countries not identified by this process. This

Supplementary Material
Supplementary Material can be found online at http://www.ntr. oxfordjournals.org

Funding
This work was supported by Cancer Research UK and the UK Department of Health.

Declaration of Interests
RW undertakes research and consultancy for companies that develop and manufacture smoking cessation medications. He also has a share of a patent for a novel nicotine delivery device and is a trustee of QUIT, a charity that provides stop-smoking support. He and SM are codirectors of the NHS Centre for Smoking Cessation and Training, a Department of Health-funded centre whose role is to identify competences necessary for behavioral support for smoking cessation and assess, and train staff in, those competences.

Acknowledgments
We would like to thank Zoe Stavri for her assistance in coding the BCTs and Peter Hajek and Andy McEwen for their feedback on the taxonomy.

References
Abraham, C., & Michie, S. (2008). A taxonomy of behavior change techniques used in interventions. Health Psychology, 27, 379387. doi:10.1037/0278-6133.27.3.379 Bauld, L., Chesterman, J., Judge, K., Pound, E., & Coleman, T. (2003). Impact of United Kingdom National Health Service smoking cessation services: Variations in outcomes in England. Tobacco Control, 12, 296301. doi:10.1136/tc.12.3.296

Nicotine & Tobacco Research


Department of Health. (1999). Smoking kills: A white paper on tobacco. London, UK: The Stationery Office. Retrieved from http://www.dh.gov.uk/en/index.htm Etter, J. F., & Stapleton, J. A. (2006). Nicotine replacement therapy for long-term smoking cessation: A meta-analysis. Tobacco Control, 15, 280285. doi:10.1136/tc.2005.015487 Ferguson, J., Bauld, L., Chesterman, J., & Judge, K. (2005). The English smoking treatment services: One-year outcomes. Addiction, 100(Suppl. 2), 5969. doi:10.1111/j.1360-0443.2005.01028.x Hajek, P. (1989). Withdrawal-oriented therapy for smokers. British Journal of Addiction, 84, 591598. doi:10.1111/j.13600443.1989.tb03474 Hajek, P. (2010). Course handbook for SCTRP course on smoking cessation. London, UK: SCTRP. Hughes, J. R., Keely, J., & Naud, S. (2004). Shape of the relapse curve and long-term abstinence among untreated smokers. Addiction, 99, 2938. doi:10.1111/j.1360-0443.2004.00540.x Judge, K., Bauld, L., Chesterman, J., & Ferguson, J. (2005). The English smoking treatment services: Short-term outcomes. Addiction, 100(Suppl. 2), 4658. doi:10.1111/j.1360-0443.2005. 01027.x McEwen, A., Hajek, P., McRobbie, H., & West, R. (2006). Manual of smoking cessation. Oxford: Blackwells. Parrott, S., Godfrey, C., Raw, M., West, R., & McNeill, A. (1998). Guidance for commissioners on the cost effectiveness of smoking cessation interventions. Health Educational Authority. Thorax, 53(Suppl. 5 Pt 2), S1S38. doi:10.1136/thx.53.2008.S2 U.S. Department of Health and Human Services. (2008). Treating Tobacco Use and Dependence: 2008 update of clinical practice guideline. Rockville, MD: Author. West, R. (2009). The multiple facets of cigarette addiction and what they mean for encouraging and helping smokers to stop. Journal of Chronic Obstructive Pulmonary Disease, 6, 277283. doi:10.1080/15412550903049181 West, R., McNeill, A., & Raw, M. (2000). Smoking cessation guidelines for health professionals: An update. Health Education Authority. Thorax, 55, 987999. doi:10.1136/thorax.55. 12.987 West, R., Walia, A., Hyder, N., Shahab, L., & Michie, S. (2010). Behavior change techniques used by the English Stop Smoking Services and their associations with short-term quit outcomes. Nicotine & Tobacco Research, 12, 742747. doi:10.1093/ntr/ ntq074

Downloaded from ntr.oxfordjournals.org at Society for Research on Nicotine and Tobacco member access on August 8, 2011

Vous aimerez peut-être aussi