Abstract

The ‘quit4u’ stop smoking service (SSS) was developed by National Health Service (NHS) Tayside for smokers in deprived areas of Dundee (UK). quit4u combined behavioural support and pharmacotherapy with financial incentives for each week that participants remained quit. A quasi-experimental study was undertaken with smokers using quit4u between 2009 and 2011 compared with smokers using SSSs in the rest of Scotland. The outcome measures were: number of quit attempts; quit rates at 1, 3 and 12 months; cost-effectiveness. Mechanisms of change were explored through quantitative and qualitative research that explored the views and experiences of service users and professionals involved in quit4u. The number of quit attempts made using SSSs in deprived areas of NHS Tayside increased by 44% between 2007 and 2010. quit4u had significantly higher quit rates at 1 month (49.9% versus 33.7%), 3 months (30.7% versus 14.2%) and 12 months (9.3% versus 6.5%) compared with similar smokers using other SSSs. The incremental cost per quitter was £2296. A combination of elements kept clients engaged and supported quit attempts: carbon monoxide (CO) tests, financial incentives, high-quality pharmacy support, rolling groups and greater varenicline use. quit4u may provide an effective and cost-effective model for engaging and supporting smokers in deprived areas to quit.

Introduction

Using financial incentives to encourage positive health-related behavioural change, including quitting smoking, is of increasing interest. Debate centres on both the effectiveness of incentives in promoting changes to complex behaviours and on the potential ‘moral hazards’ involved in doing so [1–3]. Incentives might improve the effectiveness of stop smoking services (SSSs) through increasing take-up and/or quit rates. The most recent systematic review on the effectiveness of financial incentives in smoking cessation [4] found some evidence that incentives increased participant numbers. However, the evidence for the impact on quit rates was far weaker. Only one of 19 trials [5] had significantly higher quit rates 6 months after enrolment. In other studies, early higher quit rates disappeared at longer-term (6–12 months) follow-up. Cahill and Perera [4] conclude that incentives do not enhance long-term cessation rates but could potentially increase overall numbers of quitters by increasing uptake. A review on using financial incentives to encourage healthy behaviours more generally reached similar conclusions [6]. However, a review of literature from the last 5 years [2] is more positive, suggesting that ‘when supported with other intervention strategies, such as social support and skill training, personal financial incentives can be effective in changing more complex behaviours such as reducing smoking’. This additional support may help reinforce intrinsic motivations to quit, at the same time as additional, extrinsic motivation is provided. This article aims to add to the limited evidence base by reporting the findings of the largest UK study to assess the effectiveness of combining financial incentives with free pharmacotherapy and behavioural support provided as part of the National Health Service (NHS) SSS.

The quit4u SSS in NHS Tayside combines financial incentives with behavioural support and pharmacotherapy and aims to increase uptake of cessation support and quit rates. quit4u is targeted at smokers in deprived areas. Smoking is the most important preventable cause of premature mortality and health inequalities in the United Kingdom [7]. Although there has been considerable success in reducing smoking in Britain, there has been no reduction in socioeconomic inequalities in smoking [8]. In 2011, 40% of people living in the most deprived areas of Scotland smoked compared with 10% in the least deprived areas [9]. A key element of the UK and Scottish governments’ tobacco control strategies is providing smoking cessation support through local NHS SSSs [10–12]. These services are effective in helping smokers quit and reduce inequalities by targeting disadvantaged groups [13–15]. However, quit rates among low socioeconomic status (SES) smokers who use SSSs are lower than more advantaged smokers [15]. The wider research evidence shows that, irrespective of the type of support provided, low SES smokers have lower quit rates [16]. Factors that make it more difficult for low SES smokers to quit include higher dependence levels, more difficult lives and life events which cause relapse, positive smoking social norms and lower self-efficacy [8, 17, 18]. There is therefore considerable concern to develop more effective cessation support for disadvantaged smokers.

The aim of this article is to assess the effectiveness (quit rates and number of quit attempts) and cost-effectiveness of quit4u (cost per quitter), and to identify the key factors that contributed to quit rates.

Methods

A quasi-experimental study was undertaken using routine data, comparing quit attempts made using quit4u to quit attempts made using other SSSs in Scotland. Propensity scoring, as described below, was used to reduce selection bias introduced by the non-random assignment to quit4u. The mechanisms of change were explored through additional quantitative and qualitative research to elicit the views and experiences of service users and professionals involved in quit4u.

quit4u scheme

quit4u was developed and implemented by NHS Tayside in partnership with other stakeholders, particularly Dundee City Council and the Dundee Healthy Living Initiative (DHLI). It aimed to increase SSS uptake and quit rates among smokers in deprived areas. It built on the experience of the Give It Up for Baby financial incentive scheme, previously developed by NHS Tayside to encourage disadvantaged pregnant smokers to quit [19]. The quit4u scheme was targeted at smokers living in the most deprived areas of Dundee ie DEPCATS 5, 6 and 7. DEPCAT is an area-based deprivation measure based on Carstairs scores derived from national Census data [20]. They are a composite measure based on population proportions experiencing housing overcrowding, male unemployment, low social class and having no car. The Carstairs scores are used to define seven DEPCAT groups from 1 (most affluent) to 7 (most deprived).

quit4u consisted of three elements:

  • A financial incentive of £12.50 weekly vouchers, for up to 12 weeks, to spend in ASDA (a major British supermarket chains owned by Walmart). Payments were made for each week the participant was verified as quit via a carbon monoxide (CO) test at a community pharmacy or cessation group, across a 20-week-period from first registration.

  • Behavioural support delivered in groups (run by DHLI, SSS staff or primary care practice nurses) or one-to-one (primarily by community pharmacists in their pharmacy). The behavioural support was based on evidence-based Scottish national guidelines for NHS SSSs [21]. This structured support helps clients plan and set a quit date, provides encouragement and motivation during the quitting process and helps deal with cravings and withdrawal symptoms.

  • Pharmacotherapy [i.e. nicotine replacement therapy, bupropion (Zyban) or varenicline (Champix)] provided by free prescription.

Effectiveness and cost-effectiveness of quit4u

Effectiveness was estimated by comparing quit4u quit rates with those of other SSSs in Scotland in 2009 and 2010. Data were abstracted from the NHS Scotland Information Services Division (ISD) smoking cessation dataset on all quit attempts made through Scottish SSSs from 2007 to 2010. The comparison group had access to behavioural support and pharmacotherapy (as described earlier) The effectiveness of quit4u was assessed using quit rates at 1, 3 and 12 months, with the assumption that all participants lost to follow-up had relapsed. The definition of quit attempts in these data is a modified version of the Russell Standard [22]. Attempts were identified as quits at 1 month if the individual had not smoked for the last 2 weeks, and at 3 and 12 months if, in addition to the former, they had not smoked more than five cigarettes since 1-month follow-up. The data includes CO validation at 1 month, but for a high proportion of cases (around 80%), a reading was not taken or unknown and quit rates were therefore based on self-reported smoking status. Self-reported smoking status was collected in person, by post, by telephone or by electronic means.

Propensity score regression adjustment methodology was used [23] to take account of any differences in the baseline characteristics of quit4u participants with the comparison group of Scottish SSS users. Characteristics that are related to smoking behaviours and that may vary across the two groups were included. These were age, gender, employment status, exemption from prescription charges, area-based deprivation and smoking history (number of cigarettes per day, time from waking to smoking, number of quit attempts in previous year). Appendix 1 shows the descriptive statistics for these characteristics across the two groups for the estimation sample. The propensity score was estimated using logistic regression. quit4u was only available to smokers living in postcode sectors with DEPCAT categories 5, 6 or 7; thus, only data on attempts from these deprivation categories were used to estimate propensity scores. The effect of quit4u was estimated by regressing smoking status (self-reported quitter) as a function of a dummy variable indicating whether the attempt was part of quit4u and the estimated propensity score using a log binomial model [23]. Confidence intervals and P-values are reported for the relative difference. As the effectiveness of quit4u may vary across intervention settings, the analysis was repeated by setting (pharmacy versus non-pharmacy).

Loss to follow-up is a known limitation of the ISD smoking cessation data. Moreover, the proportion of quit attempts lost to follow-up was lower in quit4u compared with other SSSs (Table I). It cannot be assessed whether differences in loss to follow-up between quit4u and the comparator group reflect the greater effectiveness of quit4u and/or simply differences in service engagement with the client and effort put in to following up participants. Through the weekly CO tests, quit4u encourages closer engagement between the service and the client and this may have resulted in lower loss to follow-up, and, unless all losses to follow-up relapse, higher quit rates. However, more effective services result in fewer relapses, and therefore, in lower losses to follow-up (assuming loss to follow-up is associated with relapse). The extent to which any differences in loss to follow-up reflect differences in service engagement may introduce a potential bias in favour of quit4u. Sensitivity analyses were therefore conducted to explore the robustness of results. Robustness was tested in three ways that attempted to reduce the difference in the proportion lost to follow-up between the quit4u and non-quit4u data. First, the comparisons were run separately for 2009 and 2010, as the differences in lost to follow-up between quit4u and other SSSs were smaller in 2009. Second, the comparison of 3- and 12-month quit rates was repeated, but excluding those missing at 1 month as the difference in loss to follow-up was largest at 1 month. Third, the comparisons were repeated using complete cases only. As loss to follow-up is likely to be correlated with relapse, complete case analysis produced a bias in favour of other SSSs compared with quit4u.

Table I.

Loss to follow-up at 1, 3 and 12 months

 Non-quit4u
 
quit4u
 
 N 
1 month 31 099 50.2 324 20.9 
3 monthsa,b 36 587 65.4 540 39.7 
12 monthsc 17 805 65.0 399 57.9 
 Non-quit4u
 
quit4u
 
 N 
1 month 31 099 50.2 324 20.9 
3 monthsa,b 36 587 65.4 540 39.7 
12 monthsc 17 805 65.0 399 57.9 

aIf failed to contact at 3 months, attempt was made to contact at 12 months.

bIncludes attempts from January 2009 to end of September 2010.

cIncludes 2009 attempts only.

Cost-effectiveness was estimated in terms of a cost per quitter from the perspective of the NHS, comparing quit4u with other SSSs in Scotland and quitting without using SSSs. All NHS interventions and pharmacotherapy used were costed using standard unit costs [24].

To explore whether quit4u was associated with an increase in quit attempts made using SSSs, time trends (2007–10) in number of quit attempts (using the same NHS smoking cessation data) in the three most deprived deciles were compared between Tayside (where quit4u was implemented) and other Scottish health boards. Reach (number of attempts as a function of the number of smokers) could not be assessed as no robust data were available on smoking prevalence by deprivation within health boards.

Mechanisms of change

A semi-structured telephone survey with a random sample of 130 quit4u participants (71% of those who provided valid contact details) was conducted within 2 weeks of registering for quit4u (Table II). This explored their smoking history, reasons for signing up for and experiences of quit4u and quitting experiences. A follow-up interview of around an hour was conducted 12 months later with all who could be re-contacted (60.8% n = 79). A purposive panel of 40 baseline survey respondents was interviewed at 3 months in greater depth about their quitting. The panel was selected to reflect the age and gender of Scottish SSS users, and included pharmacy and group support users (Table III). Follow-up in-depth interviews were conducted ∼12 months after they signed up for quit4u with 23 of the 25 panel members who still quit at 3 months. Four individual interviews with pharmacists and a focus group with staff who ran groups were conducted, which explored their views and experiences of quit4u.

Table II.

Characteristics of baseline and 12-month follow-up survey participants

 Baseline
 
12 month follow-up
 
 N N 
Gender 
    Male 50 38 32 41 
    Female 80 62 47 60 
Age 
    18–29 20 15 10 
    30–44 40 31 21 27 
    45+ 70 54 50 63 
DEPCATa 
    5 11 
    6 74 57 46 58 
    7 45 35 27 34 
Smoked in last 2 weeks at baseline 
    Yes 104 80 59 75 
    No 26 20 20 25 
 Baseline
 
12 month follow-up
 
 N N 
Gender 
    Male 50 38 32 41 
    Female 80 62 47 60 
Age 
    18–29 20 15 10 
    30–44 40 31 21 27 
    45+ 70 54 50 63 
DEPCATa 
    5 11 
    6 74 57 46 58 
    7 45 35 27 34 
Smoked in last 2 weeks at baseline 
    Yes 104 80 59 75 
    No 26 20 20 25 

aDEPCAT is an area-based measure of deprivation on Carstairs scores derived from Census data. They are a composite measure of four variables: overcrowding, male unemployment, low social class and having no car. There are seven DEPCAT groups from one (the most affluent) to seven (the most deprived) (20).

Table III.

Characteristics of 3-month qualitative panel

Age Under 30 30–44 45+ Total 
Male 10 20 
Female 20 
Signed up for pharmacy support 19 
Signed up for group support 13 21 
Quit at 3- month interviewa 12 23 
Total in each age group 13 19 40 
Age Under 30 30–44 45+ Total 
Male 10 20 
Female 20 
Signed up for pharmacy support 19 
Signed up for group support 13 21 
Quit at 3- month interviewa 12 23 
Total in each age group 13 19 40 

aThis includes some people who had relapsed and then quit again subsequently.

All those signing up for quit4u between September 2009 and February 2010 were asked if they were willing for their contact details to be passed to the research team. A study leaflet was given to potential participants. A random sample of those who opted-in was sent written information about the study, providing a further opportunity to opt out. Participants were assured of confidentiality and anonymity. Formal consent was obtained from quit4u participants and professionals (who were also provided information about the study prior to consent). Survey respondents were sent a £5 voucher to thank them for their participation. Interview participants were given £15.

The baseline and 12-months surveys were analysed using SPSS to create basic descriptive statistics (given the purpose of the surveys and the small sample sizes and relatively large confidence intervals; further, more complex analysis of this data was not deemed necessary or appropriate). The qualitative interviews and focus group were recorded and transcribed verbatim. Transcripts were analysed using a thematic approach. Transcripts were coded and summarized under key themes and sub-themes using Framework [25] and NVivo software to ensure that conclusions could be clearly linked back to original source data. Coded and summarized data were thoroughly reviewed by research team members (R.O., S.M., A.A.) to ensure the systematic mapping of the full range of attitudes and behaviour described by participants on key themes. The accounts of different participants, or groups of participants, were compared and contrasted. The research team worked collaboratively to reach consensus in interpretation and understanding. Quotations from interviews are identified by participant (P) number and smoking status at the time of their second (12-month) interview.

Ethics

Ethical approval was obtained from NHS Tayside Research Ethics Committee (IRIS) and the NatCen Social Research Ethics Committee.

Results

Number of quit attempts

quit4u signed up 2042 smokers between its launch in March 2009 and March 2011. Across Tayside as a whole, the number of quit attempts made through SSSs in the three most deprived areas increased by 44%, from 2372 in 2007/2008 to 3421 in 2009/2010. Deprivation was measured by Scottish Index of Multiple Deprivation (SIMD) rather than Carstairs DEPCAT (the measure used in quit4u) as Carstairs was not available in the 2007 and 2008 data. SIMD is highly correlated with Carstairs. quit4u was responsible for around 35.5% of all quit attempts in the three most deprived areas in Tayside in 2009 and 2010. However, other Health Board SSSs also achieved increases in quit attempts in their three most deprived areas over the same period, in response to government targets, ranging from 42% in Ayrshire and Arran to 134% in Dumfries and Galloway. This suggests that other ways of increasing uptake were available.

Quit rates

Table IV shows quit rates at 1, 3 and 12 months for quit4u attempts in 2009/2010 compared with quit rates for smokers from other SSSs, adjusted for baseline differences in age, gender, employment status, exemption from prescription charges, area-based deprivation and smoking history by including the propensity score in the regression model. Half of quit4u attempts (49.9%) were quit at 1 month, 30.7% at 3 months and 9.3% at 12 months. At each time point, the quit4u quit rate was significantly higher than that for other Scottish SSSs. Although quit rates in both quit4u and other SSSs fell sharply between 3 and 12 months, the quit4u quit rate remained significantly higher.

Table IV.

Self-reported quit rates quit4u and non-quit4u stop smoking services adjusted for baseline differences in characteristics

 Non-quit4u (other SSSs, %) quit4u (%) Relative differencea 95% CI for difference P-value N 
All  
    1-month quit rate 33.7 49.9 1.479 (1.404–1.558) 0.00 55 350 
    3-month quit rateb 14.2 30.7 2.158 (1.985–2.347) 0.00 49 887 
    12-month quit rateb 6.5 9.3 1.443 (1.132–1.839) 0.00 24 282 
Pharmacy  
    1-month quit rate 26.4 50.7 1.917 (1.784–2.060) 0.00 36 613 
    3-month quit ratec 10.2 29.3 2.857 (2.543–3.211) 0.00 33 136 
    12-month quit rateb 4.3 7.4 1.731 (1.155–2.595) 0.00 15 224 
Non-pharmacy  
    1-month quit rate 46.9 48.0 1.023 (0.950–1.102) 0.54 18 737 
    3-month quit ratec 21.5 31.2 1.446 (1.283–1.629) 0.00 16 751 
    12-month quit rateb 9.9 10.9 1.094 (0.808–1.482) 0.56 9058 
 Non-quit4u (other SSSs, %) quit4u (%) Relative differencea 95% CI for difference P-value N 
All  
    1-month quit rate 33.7 49.9 1.479 (1.404–1.558) 0.00 55 350 
    3-month quit rateb 14.2 30.7 2.158 (1.985–2.347) 0.00 49 887 
    12-month quit rateb 6.5 9.3 1.443 (1.132–1.839) 0.00 24 282 
Pharmacy  
    1-month quit rate 26.4 50.7 1.917 (1.784–2.060) 0.00 36 613 
    3-month quit ratec 10.2 29.3 2.857 (2.543–3.211) 0.00 33 136 
    12-month quit rateb 4.3 7.4 1.731 (1.155–2.595) 0.00 15 224 
Non-pharmacy  
    1-month quit rate 46.9 48.0 1.023 (0.950–1.102) 0.54 18 737 
    3-month quit ratec 21.5 31.2 1.446 (1.283–1.629) 0.00 16 751 
    12-month quit rateb 9.9 10.9 1.094 (0.808–1.482) 0.56 9058 

aEstimated by regressing smoking status (self-reported quitter) as a function of a dummy variable indicating whether the attempt is part of quit4u and the estimated propensity score using a log binomial model.

bIncludes 2009 attempts only.

cIncludes attempts from January 2009 to end of September 2010.

Pharmacotherapy was used in the majority of quit attempts (88% in quit4u and 98% in other SSSs). Use of group-based support (47% versus 18%) and varenicline (18% versus 11%) were both higher in quit4u than other SSSs. Quit rates in quit4u were higher for both pharmacy and non-pharmacy (primarily group-based) support (Table IV). The differences in quit rates were, however, much smaller for non-pharmacy and not significant at 1 and 12 months. The results of the sensitivity analyses showed that on average quit4u quit rates were higher than other SSSs quit rates with the exception of the 1-month quit rate in complete case analysis. However, the differences in quit rates were somewhat smaller and the difference in quit rates at 12 months was no longer statistically significant in two out of four scenarios.

Cost-effectiveness

The quit4u costs included the financial incentives, administrative costs and other service costs including pharmacotherapy. The average total cost per attempt was estimated at £191 for quit4u and £98 for other SSSs. In comparison with self-quit (assuming 1% annual quit rate [26]) and taking into account intervention costs only, quit4u cost an additional £2296 per 12-month quitter whilst other SSSs on average cost an additional £1797 per 12-month quitter.

Mechanisms of change—joining quit4u

Participants’ reasons for joining quit4u were explored in the baseline survey and qualitative interviews (at 3 months). Four main categories of reasons for signing up were identified in the qualitative interviews: the financial incentive, pharmacotherapy (varenicline and nicotine patches were specifically mentioned), structured support in general, or to receive specific support such as group therapy or CO breath tests. Thus all the quit4u elements appeared to have played a role in attracting different participants. These categories were not mutually exclusive. For example, participants described being attracted by both the financial incentive and the structured support.

Participants’ perceptions of the role of financial incentives in encouraging sign-up fell into four groups: those who described the incentive as the main reason for signing up to quit4u, those who regarded the incentive as an important factor but not the main reason, those who stated that the incentive was not a reason for joining but a bonus and those who said it had not been a factor at all. In the two latter groups the desire to get help to quit smoking was given as the main reason, with any possible financial gain being secondary or irrelevant.

Participants’ claims that they were not motivated by money may have been influenced by social desirability bias. However, among those who insisted that the incentives had not been a factor were participants who stated they had not known about the incentive at the time they decided to sign up. In addition, several participants stated that they had not claimed the incentive and did not intend to or would give the money away.

That the incentives did not dominate thinking about joining quit4u was indicated by over twice as many baseline survey respondents (87% n = 113) stating that they thought the CO breath tests would help motivate them to stay quit compared with being motivated to sign-up for the scheme by the incentive (36% n = 47).

Participants who felt that the incentives had been a primary or a secondary reason for their signing-up indicated that it provided a trigger or ‘tipping point’ to give up and to give up with support. These views were echoed by community pharmacists and smoking cessation advisors.

Mechanisms of change—sustaining quit attempts

Participants’ perceptions of whether and in what ways quit4u had helped sustain their quit attempt were explored in the follow-up survey (at 12 months) and qualitative interviews (at 3 months). Most survey respondents (77% n = 61) reported that they had found trying to quit with quit4u easier than previous attempts. Qualitative interview participants discussed the ways in which they felt this quit attempt had differed from previous attempts. In particular they discussed their perceptions of quit4u as a whole and the individual elements which they felt had contributed to their success or otherwise in quitting.

The ‘whole package’

There was no consensus among survey respondents or interviewees over what was the most helpful element of quit4u. Although different individual components were singled out by individuals, there was also a view that it was ‘the whole package’, or the overall ‘structure’ of quit4u which had helped. This view was shared by successful and unsuccessful quitters.

‘All these things come together, don’t they? They become part of the whole ‘stopping smoking', because you’re going there once a week for 12 weeks or something.’ (P172, quit)

The financial incentive

Of the 62 survey respondents who had received payments, 71% (n = 44) felt these had been ‘very’ or ‘quite useful’ in helping them stay quit. Qualitative interview participants were split between those who believed the incentive had been the main factor in their staying quit, those who felt it had been a secondary or additional factor compared with their determination to quit or other service elements (e.g. behavioural support, pharmacotherapy, CO test), and those who maintained the incentives had not been a factor in continuing with quit4u or maintaining their quit attempt. These views did not appear to be related to quit success.

Among those who felt it had been a factor in encouraging them to stay quit, the incentive was viewed as: a ‘reward’, ‘bonus’ or ‘wee treat’ for quitting; a ‘wee bit extra to keep you going’ with the quit attempt; providing ‘something to work towards’; and a reason to stick with the programme or to keep going back to the pharmacist.

‘I was going to stop anyways. (It was) just a, certainly a bonus having that there like. I reckon it would probably encourage you to keep going back to the chemist.’ (P209, quit)

Similar views were expressed by service providers, who suggested that the incentive encouraged participants to stick with formal support for longer, which in turn increased their chances of quitting successfully. It was also suggested that the financial incentive might encourage participants to come back for support, even if they relapsed.

‘It’s also an incentive to come back as well to the Pharmacy, even if they have sort of failed in the first couple of weeks – again, you just use that to actually reinforce that “OK, it’s difficult to start with, but there’s also £12.50 a week if you do sort of come a bit better towards the end.’ (Pharmacist 1)

However, none of the survey respondents identified the incentive as being the most useful component of quit4u. Pharmacotherapy, behavioural support, CO tests or a combination of these were identified as more important in helping quit initially and stay quit. This was again reflected in the views of providers, who suggested that the financial incentive in itself was not the key to quitting successfully, but that it helped keep more people engaged with support (pharmacological and personal) for longer.

CO tests

The weekly CO tests were regarded by participants as a novel and important part of quit4u. Participants described how the test not only validated their quit status, but also provided an additional reason for not relapsing (not wanting to fail a test). For some, this was related to wanting to pass the tests to receive the financial incentive. For others, passing the tests appeared to provide a motivation to continue with the quit attempt in itself.

‘It was good getting the breathalysers … That made my day for some reason! What I wouldn’t say was a great factor, was the money side of things. That didn’t really, it didn’t help me.’ (P212, quit)

‘I mean on the scheme that I was on before … there was nothing like that, so you could go in and go “Oh, no. No. I've not been smoking”. But with this, you will know. So that’s another sort of incentive to go “Oh, I’d better not, ‘cos as soon as I go and blow in that they're going to know!” So that’s quite a good thing that.’ (P159, relapsed)

The tests were also viewed as useful, by participants and providers, in demonstrating immediate health benefits from quitting.

‘I mean it made me feel good that. I kind of had pictures in my head of my lungs getting better.’ (P142, relapsed)

Participants described looking forward to the test, a reason to attend the weekly support sessions. In groups, the tests were also described as providing a competitive element, which could be an incentive to continue with the quit.

Behavioural support

For participants who attended groups, being able to share quitting experiences with fellow quitters, as well as getting advice from group leaders and members, was considered helpful and motivating. quit4u had ‘rolling’ groups where new people joined at different points. These were seen as more beneficial than the traditional model of groups where everyone starts together, because they included people who had been through the stage of quitting that other participants were experiencing and with whom could share their experiences, advice and tips.

‘It was handy just to pop to the Community Centre down the road, but also listen to different suggestions from different people all of what worked and also maybe passing that information on to other new ones that came a few weeks down the line as well. I found that obviously good, that you were able to help other people if you had information.’ (P32, quit)

Seeing the success of other group members, as well as providing encouragement and support to those at an earlier quitting stage, was viewed as motivational by participants and providers. Group leaders were also praised for their non-judgemental and supportive attitudes.

Although participants’ accounts of the level and quality of the one-to-one support from pharmacy staff varied, these included some very positive comments about the support received. In some cases, this was contrasted with previous experiences where participants reported that they had just been given their pharmacotherapy and perhaps a leaflet, with little behavioural support or encouragement.

‘This time it was a lot better with going in and having the 10 minute talk to her and that about it. The last time, it was just handing over the counter and away.’ (P167, relapsed)

Participants’ accounts suggested that the CO test may have helped provide a focus for encouragement, praise and support.

‘She made it good to go in there and you know, breath into your wee machine … And she got just as excited as you when it was just on the little ‘1’ thing.’ (P88 relapsed)

Participants commented on the positive and interested attitude of pharmacy staff, their level of knowledge (e.g. helping if there were problems with the pharmacotherapy such as side-effects) and the time given to explaining pharmacotherapy options, encouraging participants and talking to them about how they were getting on.

Pharmacotherapy

Most participants in the qualitative and survey research had used pharmacotherapy in previous quit attempts. Thus using pharmacotherapy was not generally cited as something that made quit4u different from previous quit attempts, except where participants had not used pharmacotherapy previously or felt that the pharmacotherapy they had used this time was more effective (e.g. varenicline). It was, however, suggested that quit4u had provided participants with more information and choice about products, and that they were able to go back to groups or pharmacies if they had problems for advice or changes in medication.

‘It wasn’t a case of “Oh, try this, this if for you”. It was “try this, if it doesn’t work for you come back and we’ll try something different”.’ (P86, relapsed)

Discussion

This is the largest study in the United Kingdom to have evaluated including financial incentives as part of smoking cessation support. It focused on disadvantaged smokers who have the lowest quit rates. The study used a range of quantitative and qualitative data and methods to assess outcomes and explore the reasons for these. The results showed that quit4u had higher quit rates at 1, 3 and 12 months than other SSSs in Scotland, after adjusting for differences in participants’ baseline characteristics, such as deprivation and smoking dependence. The average cost per quitter was relatively low (£191) and given the reduced NHS treatment cost of smoking related diseases associated with quits [26, 27], a highly cost-effective use of resources.

Effectiveness was assessed using a quasi-experimental design rather than a randomized controlled trial design, which has implications for the robustness of the evidence. Non-random assignment can result in a selection bias. Although a quasi-experimental design is more likely to suffer from potential biases, a strength was that the service was delivered under ‘real life’ conditions. Propensity scoring was used to reduce selection bias from non-random assignment. However, the two groups may have also differed in terms of unobserved characteristics, such as income, which the analysis could not control for. There were also concerns over the quality of the data in terms of extent of and differences in loss to follow-up. The higher number of missing data in the comparison group may have underestimated the effectiveness of the comparison group and introduced a bias in favour of quit4u. Sensitivity analyses were conducted to explore the impact of differences in loss to follow-up on results. Differences in quit rates were lower in the sensitivity analyses and in two out of the four scenarios the difference in 12-month quit rates was no longer significant. This suggests that there is uncertainty around the longer-term effectiveness. The sensitivity analysis included complete case analysis that would usually not be recommended with a relatively high proportion of missing data. Finally, the outcome measure was based on self-reported smoking status. Evidence suggests that the use of self-reported smoking status leads to over-estimation of quit rates [28]. The interest in this article is in differences in quit rates. The difference in rates will only be biased, if the tendency to over-estimate varies across quit4u attempts and other attempts. Although it could be argued that this may be the case for the 1 and 3-month outcomes as quit4u participants knew that their smoking status will be verified by CO tests, it is less clear why a difference would exist in the case of the 12-month outcome. It is also possible that taking part in the quit4u evaluation survey and interviews may have impacted on participants’ quit success. However, only 130 out of 2042 quit4u clients took part in the baseline survey and therefore any impact on quitting success is unlikely to have been significant.

quit4u’s higher quit rates appear to be mainly due to the significantly higher quit rates achieved through one-to-one pharmacy support, particularly at 1 and 3 months. Participants’ accounts of the support received from pharmacies suggest that CO tests may have given an additional focus for providing encouragement and support which may, in turn, have improved engagement between pharmacy staff and clients. quit4u’s higher use of cessation groups and varenicline are also likely to have contributed to the higher quit rates, but to a lesser extent. Previous research has indicated that group interventions are more effective in supporting quitting than one-to-one support [13, 21, 29] and varenicline is more effective than other types of pharmacotherapy [21]. quit4u quit rates at 3 and 12 months were higher for those who used group support, and higher at 1 and 3 months for those who used varenicline. The comparatively lower level of loss to follow-up in quit4u suggests that a key reason for its higher quit rates was its greater success in maintaining contact with service users. Previous studies have found that low SES smokers drop out of services earlier than higher SES smokers [8, 30]. There was no consensus among quit4u participants over which element was most helpful in sustaining their quit attempt. Participants who felt that the financial incentive had helped them maintain their quit attempt for longer saw it as providing a bonus, something to work towards and an encouragement to keep coming back for support. The perceived role of incentives in encouraging participants to stick with support for longer was also highlighted by service providers.

The CO tests gave participants an additional reason not to smoke, which appeared to go beyond the desire to ‘pass’ in order to claim the financial incentive. The tests were viewed as demonstrating the immediate health benefits of quitting, providing an element of competition, positive reinforcement or reward and creating additional motivation not to fail. These experiences were similar to those found in a pilot study, which used financial incentives as part of stop smoking support for pregnant smokers [31]. Women who were incentivized reported greater engagement with the service and described the motivating experience of being CO monitored and getting feedback on their progress. quit4u’s use of rolling groups was also viewed positively. As has been found in other studies [32, 33], being able to share the experiences, successes and tips of participants at different stages of their ‘quit journeys’ was motivational. Thus quit4u would appear to increase extrinsic and possibly intrinsic motivation, leading to higher engagement and quit rates at 1 and 3 months. The relative decline in quit rates by 12 months indicates that any increase in intrinsic motivation may not be sustained.

It was not possible to isolate or quantify the unique or additional contribution of the incentives. However, this should not be regarded as a weakness of the evaluation. Attempting to separate out the effect of the incentive would have been inappropriate and misleading as other inherently linked elements, notably the CO tests, were also key. Mantzari et al. [31] have argued (2012) ‘we need to be cautious about attributing the effects of financial incentives schemes to incentives per se’ as they might operate through various pathways.

Previous research has suggested that financial incentives might increase the effectiveness of smoking cessation support by increasing the take-up of services and their quit rates. quit4u increased the number of disadvantaged smokers registering for cessation support. However, other Health Board SSSs achieved increases in take-up among smokers in deprived areas over the same period. This probably reflected increased activity by Scottish SSSs to achieve the national Health Improvement, Efficiency, Access and Treatment targets by the March 2011 deadline (to support 8% of their smoking population to quit, measured at 4 weeks post-quit date, between 2008 and 2011, with priority given to deprived areas [12]. Thus it is not possible, based on the available data, to determine whether quit4u was more successful in increasing take-up of cessation services than approaches in other Health Board areas. Although financial incentives appeared to provide a trigger or ‘tipping point’ to quitting and getting support for a substantial minority of participants, so did the other elements of quit4u including the CO tests, support and pharmacotherapy.

In conclusion, this evaluation indicates that quit4u may provide an effective and cost-effective model for engaging and supporting smokers in deprived areas to quit. The quantitative and qualitative data provide important insights about the elements of quit4u that contributed to the effectiveness of this model in keeping clients engaged and supporting them to quit successfully. These include CO tests, high-quality pharmacy support, rolling group support, greater use of varenicline and financial incentives. In combination, these elements appear to provide an effective model for engaging and supporting smokers in deprived areas to quit.

Acknowledgements

The authors thank all those involved in quit4u-NHS Tayside, partners and participants. The authors particularly like to thank Fiona Myers (NHS Health Scotland), Andrew Radley (NHS Tayside), Paul Ballard (NHS Tayside) and Linsey Galbraith (ISD).

Funding

This work was supported by NHS Health Scotland. The Chief Scientist Office of the Scottish Government Health Directorates funds HERU. The opinions expressed in this paper are those of the authors and are not necessarily those of the NHS Health Scotland.

Conflict of interest statement

None declared.

References

1
Marteau
T
Ashcroft
R
Oliver
A
Using financial incentives to achieve healthy behaviour
BMJ
 , 
2009
, vol. 
338
 pg. 
b1415
 
2
Lynagh
M
Sanson-Fisher
R
Bonevski
B
What’s good for the goose is good for the gander: guiding principles for the use of financial incentives in health behaviour change
Int J Behav Med
 , 
2013
, vol. 
20
 (pg. 
114
-
20
)
3
NICE Citizens Council
The Use of Incentives to Improve Health. Citizens Council Report 14
 , 
2010
London
NICE
 
4
Cahill
K
Perera
R
Competitions and Incentives for Smoking Cessation
Cochrane Database Syst Rev
 , 
2008
, vol. 
3
  
: Art. No.: CD004307
5
Volpp
KG
Troxel
AB
Pauly
MV
, et al.  . 
A randomized, controlled trial of financial incentives for smoking cessation
N Engl J Med
 , 
2009
, vol. 
360
 (pg. 
699
-
709
)
6
Jochelson
K
Paying the Patient: Improving Health Using Financial Incentives
 , 
2007
London
King’s Fund
7
Jha
P
Peto
R
Zatonowksi
W
, et al.  . 
Social inequalities in male mortality, and in male mortality from smoking: indirect estimation from national death rates in England and Wales, Poland, and North America
Lancet
 , 
2006
, vol. 
368
 (pg. 
367
-
70
)
8
Hiscock
R
Judge
K
Bauld
L
Social inequalities in quitting smoking: what factors mediate the relationship between socioeconomic position and smoking cessation?
J Public Health
 , 
2010
, vol. 
33
 (pg. 
39
-
47
)
9
Scottish Government
Scotland's People Annual Report: Results from 2011 Scottish Household Survey
 , 
2012
Edinburgh
National Statistics/Scottish Government
10
Department of Health
A Smokefree Future—a Comprehensive Tobacco Control Strategy for England
 , 
2010
London
Department of Health
11
Department of Health
Healthy Lives, Healthy People: A Tobacco Control Plan for England
 , 
2011
London
Department of Health
12
Scottish Government
Creating a Tobacco-Free Generation—A Tobacco Control Strategy for Scotland
 , 
2013
Edinburgh
Scottish Government
13
Bauld
L
Bell
K
McCullough
L
, et al.  . 
The effectiveness of NHS smoking cessation services: a systematic review
J Public Health
 , 
2010
, vol. 
32
 (pg. 
71
-
82
)
14
Bell
K
McCullough
L
Greaves
L
, et al.  . 
NICE Rapid Review—The Effectiveness of NHS Intensive Treatments for Smoking Cessation in England
 , 
2006
London
NICE
15
Galbraith
L
Hecht
G
NHS Smoking Cessation Service Statistics (Scotland) 1st January to 31st December 2010
 , 
2011
Edinburgh
ScotPHO
16
Amos
A
Bauld
L
Clifford
D
, et al.  . 
Tobacco Control, Inequalities in Health and Action at a Local Level
 , 
2011
York
Public Health Research Consortium
17
Hiscock
R
Bauld
L
Amos
A
, et al.  . 
Smoking and socioeconomic status in England: the rise of the never smoker and the disadvantaged smoker
J Public Health
 , 
2012
, vol. 
34
 (pg. 
390
-
6
)
18
Fidler
JA
Jarvis
MJ
Mindell
J
, et al.  . 
Nicotine intake in cigarette smokers in England: distribution and demographic correlates
Cancer Epidemiol Biomarkers Prev
 , 
2008
, vol. 
17
 (pg. 
3331
-
6
)
19
Ballard
P
Radley
A
Give It Up for Baby—a smoking cessation intervention for pregnant women in Scotland
Cases Public Health Commun Market
 , 
2009
, vol. 
3
 (pg. 
147
-
60
)
20
McLoone
P
Carstairs Scores for Scottish Postcode Sectors from the 2001 Census
 , 
2004
Glasgow
MRC Social and Public Health Sciences Unit
 
21
NHS Health Scotland
A Guide to Smoking Cessation in Scotland 2010—Planning and Providing Specialist Smoking Cessation Services
 , 
2010
Edinburgh
NHS Health Scotland
22
West
R
Assessing Smoking Cessation Performance in NHS Stop Smoking Services: The Russell Standard (Clinical)
 , 
2005
London
Cancer Research UK/UCL
23
Austin
PC
An introduction to propensity-score methods for reducing confounding in observational studies
Multivariate Behav Res
 , 
2011
, vol. 
46
 (pg. 
399
-
424
)
24
NICE
Smoking Cessation Services: Costing Template
 , 
2009
London
NICE
 
25
Ritchie
J
Spencer
L
Bryman
A
Burgess
R
Qualitative data analysis for applied policy research
Analysing Qualitative Data
 , 
1993
London
Routledge
26
Godfrey
C
Parrott
S
Coleman
T
, et al.  . 
The cost-effectiveness of the English smoking treatment services: evidence from practice
Addiction
 , 
2005
, vol. 
100
 
Suppl. 2
(pg. 
70
-
83
)
27
Godfrey
C
Ali
S
Parrott
S
Pickett
K
Economics Model of Adult Smoking Related Costs and Consequences for England
 , 
2011
York
Public Health Research Consortium
28
Connor Gorber
S
Schofield-Hurwitz
S
Hardt
J
, et al.  . 
The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine assessed smoking status
Nicotine Tob Res
 , 
2009
, vol. 
11
 (pg. 
12
-
24
)
29
Brose
L
West
R
McDermott
M
, et al.  . 
What makes for an effective stop smoking service?
Thorax
 , 
2011
, vol. 
66
 (pg. 
924
-
6
)
30
Kotz
D
West
R
Explaining the social gradient in smoking cessation: it’s not in the trying, but in the succeeding
Tob Control
 , 
2009
, vol. 
18
 (pg. 
43
-
6
)
31
Mantzari
E
Vogt
F
Marteau
TM
The effectiveness of financial incentives for smoking cessation during pregnancy: is it from being paid or from the extra aid?
BMC Pregnancy Childbirth
 , 
2012
, vol. 
12
 pg. 
24
 
32
Ritchie
D
Schulz
S
Bryce
A
One size fits all? A process evaluation—the turn of the ‘story’ in smoking cessation
Public Health
 , 
2007
, vol. 
121
 (pg. 
341
-
8
)
33
Springett
J
Owens
C
Callaghan
J
The challenge of combining lay knowledge with evidence-based practice in health promotion: Fag ends smoking cessation service
Critical Public Health
 , 
2007
, vol. 
17
 (pg. 
243
-
56
)

Appendix

Table A1:

Characteristics of quit attempts (non-quit4u SSSs and quit4u)

 Non-quit4u (other SSSs)
 
quit4u
 
χ2 (P-value) 
 N N  
Gender 
    Female 30 790 61.5 645 62.9 0.839 (0.36) 
    Male 19 308 38.5 381 37.1 
Age 
    <18 851 1.7 0.6 28.56 (0.01) 
    18–24 3660 7.3 68 6.6 
    25–34 9162 18.3 138 13.5 
    35–44 11 978 23.9 256 25.0 
    45–59 15 776 31.5 369 36.0 
    >60 8671 17.3 189 18.4 
Employment status 
    Fulltime student 1872 3.7 33 3.2 34.03 (0.01) 
    Homemaker/fulltime parent or carer 2563 5.1 38 3.7 
    Paid employment 20 623 41.2 444 43.3 
    Permanently sick/disabled 3864 7.7 105 10.2 
    Retired 7570 15.1 166 16.2 
    Unemployed 11 463 22.9 225 21.9 
    Other 2143 4.3 15 1.5 
Ethnicity 
    White 49 672 99.1 1023 99.7 3.76 (0.05) 
    Non-white 426 0.9 0.3 
Free prescription 
    No 13 085 26.1 314 30.6 10.46 (0.01) 
    Yes 37 013 73.9 712 69.4 
Carstairs deprivation category 
    5 20 103 40.1 171 16.7 277.34 (0.01) 
    6 16 685 33.3 561 54.7 
    7 13 310 26.6 294 28.7 
Smoking history 
Cigarettes per day 
    <10 5661 11.3 104 10.1 22.98 (0.01) 
    11–20 22 569 45.0 539 52.5 
    20–30 14 735 29.4 261 25.4 
    >30 7133 14.2 122 11.9 
How soon after waking 
    Within 5 min 29 056 58.0 559 54.5 13.11 (0.01) 
    6–30 min 15 161 30.3 363 35.4 
    31–60 min 3641 7.3 66 6.4 
    Over 1 h 2240 4.5 38 3.7 
How many times tried to quit in the past year 
    Never 21 205 42.3 391 38.1 24.88 (0.01) 
    Once 16 693 33.3 318 31.0 
    2 or 3 times 9800 19.6 247 24.1 
    4 or more times 2400 4.8 70 6.8 
 Non-quit4u (other SSSs)
 
quit4u
 
χ2 (P-value) 
 N N  
Gender 
    Female 30 790 61.5 645 62.9 0.839 (0.36) 
    Male 19 308 38.5 381 37.1 
Age 
    <18 851 1.7 0.6 28.56 (0.01) 
    18–24 3660 7.3 68 6.6 
    25–34 9162 18.3 138 13.5 
    35–44 11 978 23.9 256 25.0 
    45–59 15 776 31.5 369 36.0 
    >60 8671 17.3 189 18.4 
Employment status 
    Fulltime student 1872 3.7 33 3.2 34.03 (0.01) 
    Homemaker/fulltime parent or carer 2563 5.1 38 3.7 
    Paid employment 20 623 41.2 444 43.3 
    Permanently sick/disabled 3864 7.7 105 10.2 
    Retired 7570 15.1 166 16.2 
    Unemployed 11 463 22.9 225 21.9 
    Other 2143 4.3 15 1.5 
Ethnicity 
    White 49 672 99.1 1023 99.7 3.76 (0.05) 
    Non-white 426 0.9 0.3 
Free prescription 
    No 13 085 26.1 314 30.6 10.46 (0.01) 
    Yes 37 013 73.9 712 69.4 
Carstairs deprivation category 
    5 20 103 40.1 171 16.7 277.34 (0.01) 
    6 16 685 33.3 561 54.7 
    7 13 310 26.6 294 28.7 
Smoking history 
Cigarettes per day 
    <10 5661 11.3 104 10.1 22.98 (0.01) 
    11–20 22 569 45.0 539 52.5 
    20–30 14 735 29.4 261 25.4 
    >30 7133 14.2 122 11.9 
How soon after waking 
    Within 5 min 29 056 58.0 559 54.5 13.11 (0.01) 
    6–30 min 15 161 30.3 363 35.4 
    31–60 min 3641 7.3 66 6.4 
    Over 1 h 2240 4.5 38 3.7 
How many times tried to quit in the past year 
    Never 21 205 42.3 391 38.1 24.88 (0.01) 
    Once 16 693 33.3 318 31.0 
    2 or 3 times 9800 19.6 247 24.1 
    4 or more times 2400 4.8 70 6.8 

Comments

0 Comments