The WERO group stop smoking competition: main outcomes of a pre- and post- study

In order to reach the goal of Smokefree 2025, the New Zealand Ministry of Health’s key approach is to trigger more supported quit attempts. To date, as far as we can tell, Quit and Win competitions have not been adapted to include cultural elements, such as focusing on family or community groups and invoking existing inter-tribal rivalry to increase the attractiveness or relevancy for different ethnic groups [35]. By increasing attractiveness and thereby enrolment in cessation interventions it is possible to overcome some of the barriers to cessation, such as lack of knowledge and low attractiveness of cessation products. Furthermore, previous studies have identified the need for social support in Quit and Win contests, however, those studies incorporated a non-contestant to support a contestant. The previous studies that incorporated the support into a team contest are over 20 years old and were only tested in a work environment. The current uncontrolled pre and post study differs from those studies in that it aims to investigate if WERO (WERO in the Māori language means challenge), a stop smoking contest, using both within team support, external support from a team coach and cessation experts, and technology, would be effective in prompting quitting and increasing abstinence from smoking.

WERO utilises several behaviour change strategies: incentives, competition, social support, behavioural therapy, pharmacological therapy and an interactive website and iPad application (app). Teams competed against each other for three months (31^st May- 30^th August 2012) to win a NZ$5000 (about €3,000, £2600, US$4000) prize to go to a charity or community group of their choice. Each team had to appoint their own coach (a non-smoker or ex-smoker) from their community. WERO co-ordinators assisted coaches to support their teams to stop smoking, to access cessation workers and pharmacotherapy and co-ordinators found in-kind support or spot prizes from within the community, such as free access to the public swimming pool, or retail vouchers from a sporting goods store. All WERO co-ordinators received training in how WERO works, their role, how to inform participants, obtain consent and collect data including exhaled carbon monoxide (CO) readings.

WERO teams were recruited via community and health promotion workers from three organisations: Northland District Health Board (who recruited five rurally based Māori teams), Te Whānau o Waipareira, a Māori health and social services provider, in West Auckland, NZ’s biggest city (who recruited five urban Māori teams), and a Pacific Island churches healthy lifestyles programme, Enua Ola, in Auckland (who recruited five Pacific Island teams). Some teams included participants of mixed and other ethnicities. The cluster of teams are hereto forth referred to as Rural Māori, Urban Māori and Pacific. For each cluster, WERO co-ordinators were appointed to recruit teams from their community networks over six weeks (from mid-April until 31^st May of 2012). In Northland, this involved the co-ordinator approaching individuals in community groups such as sports clubs, kapa haka groups, marae (traditional Māori meeting locations) and small communities. For Pacific, the WERO co-ordinator approached key members of the church communities to recruit teams. Urban Māori teams were formed from staff within different business units within Te Whānau o Waipereira, a church group, a mums and babies group and a family group.

People were eligible to participate if they were aged 18 years and over, identified as smokers and had an exhaled CO rating greater than 6 parts per million (ppm), measured with a Bedfont Smokerlyser. The intervention was focused on Māori and Pacific Island people. However, it was a desired criterion, not an inclusion criteria per se.

People who could not provide written consent, non-smokers or smokers whose exhaled CO was less than or equal to 6 ppm at entry were excluded from the study.

Fifteen teams of 10 were to be recruited. Assuming 80% power, 5% significance, 1-sided test, no design effect, and a standard treatment 3 month quit rate of 24% (from the National Quitline), the study needed approximately 100 people to detect a target 3 month quit-rate of 35%. However, because the drop-out rate at follow-up is assumed to be high (50% from past studies), 150 smokers were recruited. For one team, two contestants stopped smoking prior to completing the eligibility screening and were hence deemed ineligible. Therefore, there were 14 teams with 10 contestants and one team with eight contestants, a total of 148 participants.

At the end of competition and at follow-up, six months after the start of competition, contestants were asked “In the last 7 days have you smoked a cigarette (even a puff)?” However, interventions that use competitions or incentives suffer from high levels of deception [21]. Therefore, smoking status was biochemically validated at competition entry to determine eligibility, and at the end of the competition at 3 months to verify self-reported smoking status. Biochemical validation of smokefree status was undertaken using exhaled CO testing. For a contestant at the end of competition to be deemed to have quit smoking they had to have answered ‘No’ to the question “In the last 7 days have you smoked a cigarette (even a puff)?” and have an CO reading equal to or below 6 ppm (MacLaren et al. [36]). CO testing was chosen for pragmatic reasons, namely its low cost, ease of use and its availability for healthcare providers.

In addition, participants self-completed questionnaires, containing questions about gender, ethnicity, age, marital status and education level. Also to get feedback on the competition the following questions were included: In your opinion, what makes WERO successful? Tick box options were: Getting support to stop smoking; Getting to raise money for a good cause; Being in a competition; Stopping smoking in a team; and a free text option. Lastly, participants were asked to write about what did not work so well with the WERO competition and what could be done to improve it. The qualitative work adheres to the qualitative research reviews guidelines.

Lastly, qualitative interviews were conducted with co-ordinators and one coach from each cluster of teams at the completion of the 3 month data collection. The interviews included questions on intervention implementation, acceptability and satisfaction with the intervention, identification of barriers for effective implementation, and suggestions for improvement.

The raw data was entered into an EXCEL spreadsheet and exported to SAS v 9.2 (Cary NY). Intention to treat analysis was used for all quit-rate calculations, that is, those lost to follow-up were assumed to be smoking. To determine the baseline predictors of those who quit at 6 months we used general estimating equation (GEE) analysis to account for correlation within teams and added a region variable (i.e., 5 teams each were nested within 3 clusters) into the model statement. Our model included ethnicity, gender, age (less than or equal to 26 years versus over 26 years), time to first cigarette (less than or equal to 30 minutes versus more than 30 minutes), and region. We chose the exchange structure (also known as compound symmetry) for the correlation matrix because it had a lower QIC score (Quasi-likelihood Information Criterion) than other structures.

The qualitative data was deductively analysed. Summary data were calculated for the participants’ quantitative interview questions.

Ethical approval was obtained through the Northern X Regional Ethics Committee, reference number NTX/11/EXP/308. Due to the public nature of the contest, participants smoking status was not kept confidential. However, all other data collected were confidential. All data were analysed anonymously. Informed consent was administered by the regional co-ordinators.

The age range of participants was 16 to 70 years old, with a mean of 38 years, 53% were Māori and nearly 60% were women (Table 1). On average contestants smoked 15.6 cigarettes per day. Many (71%) had their first cigarette within 30 minutes of waking up, indicating a high level of addiction. Thirty-seven per cent had never tried to quit before.

At the end of competition the biochemically verified 24-hour quit rate was 36% (95% CI: 28%-44%) (Table 2). The 6 months self-reported quit rate was 26% (95% CI: 20%-34%). At competition end, loss to follow-up was 14% (21/148) and 19% (28/148) at 6 months.

Those aged 26 years and older were significantly more likely (OR = 1.99, 95% CI 1.08-3.64) to self-report being quit smoking at 6 months follow-up compared to younger participants (Table 3).

The urban Māori teams had more participants that were 26 years old or younger, than the rural Māori or Pacific teams (Table 4). Pacific had a high rate (80%) of smokers who had never quit before, compared to 19% and 12% for the Māori urban and rural teams.

The number of quitters per team ranged from zero (no quitters) to ten (all quit). Three teams had no quitters and one team had ten quitters. The average number of quitters per team was 3.5 and the median number of quitters per team was 3. The Pacific teams had the highest overall quit rate at both end of competition and at 6-months follow-up (Table 5). However, the winning team, with 10 out of 10 quitters, were a rural Māori team. The urban Māori teams had substantially lower quit rates at both end of competition and 6-months follow-up than the rural Māori or Pacific teams.

All coaches and co-ordinators expressed great satisfaction and acceptability of the intervention, including a coach from and the co-ordinator of the Māori urban teams that did not have as many quitters.

It was identified that the concept of WERO could be applicable in other areas also.

A number of contestants provided very positive feedback.

The most common activities that the contestants found helpful were support, from each other, the team coach and WERO co-ordinators, and getting together and competing against other teams (Table 6a). The rural Māori teams particularly found the within team support and support from coaches and co-ordinators helpful, while more of the Pacific teams selected getting together as helpful. Only 7% of participants didn’t take part in the activities, nearly all from the urban Māori teams, and only 2% did not find any of the activities helpful, all from the urban Māori teams.

Over 90% of participants selected one of more options as successful, 27% of participants did not select any options as successful and three participants stated that nothing was successful with WERO (Table 6b). The most commonly selected options were getting support to stop smoking and stopping smoking in a team.

Coaches, co-ordinators and contestants (Table 6a) identified that between-teams contest was most important.

Competing against teams from other clusters and Māori versus Pacific was also important:

Only 31 contestants provided any criticism of WERO. The most common critiques of WERO, from both coaches, co-ordinators and contestants was difficulty in getting team members together. Some contestants reported that there was lack of commitment (n = 4) or support (n = 4) from team members, and difficulty in attending meetings (n = 3). Coaches and co-ordinators also stated difficulties getting some teams together.

Another criticism from coaches was the limited time prior to the start of the contest for coaches to prepare.

An Urban Māori coach thought that the contest wasn’t long enough, while the rest of the coaches and co-ordinators thought that the length was just right.

There were also some technical difficulties. Three contestants had no access to technology (n = 3). Some teams didn’t have internet access.

There were also some concerns about the CO testing:

Only 31 participants provided any suggested improvements. The most common suggestion (n = 11) was for more prizes, for example, some prize for all who managed to get smokefree, smaller prizes at regular intervals of the competition and gifts for all participants.

Contestants, co-ordinators and coaches suggested more activities, within teams, with WERO support staff and with other teams.

The main strength of this study was that the intervention was informed by the needs of the target audience, identified in previous research by the authors. The intervention was innovative in its use of new technology (e.g. apps) applied at a community level. This fitted with the New Zealand Tobacco Control Research Tūranga brief for research to be innovative, pragmatic, able to demonstrate a strong link to rapid prevalence reduction, be relatively short in duration from conception to delivering of results and have the potential to be scaled up. The requirement to conduct shorter studies imposes some limitations. Strongly powered randomised controlled trials, for instance, are beyond the scope and budget available to the Turanga. Hence the short-term pre and post study non-randomised design used here. One limitation of the research design, is that it was not possible to control for confounders, such as the impact of individual co-ordinators or coaches. The design also limits the comparability of WERO to previous Quit and Win contests [19] and previous workplace studies [19, 22‐24], since those studies were conducted in different countries and with more rigorous design. Our study also lacked a control group so it is difficult to ascertain how much of the high quit rate was due to the external environment rather than the intervention itself. However, as far as we are aware, there were no major significant tobacco control policies, such as tax increases, during the intervention period. End-of-competition smoking status in this study was ascertained using CO tests. Given that CO levels only assess smoking in the preceding nine hours, it is possible that the quit rates reported in this study are an overestimate. Cotinine tests would have provided a more reliable assessment of smoking cessation, however, it was not feasible in this study due to costs. Lastly, 6-month follow-up relied on self-report and was not biochemically verified. This could also potentially overestimate the quit rate.

The limitations of this study highlight the need for future controlled trials with longer follow-up times of team based cessation contests like WERO. The survey and interview data suggests that the group element and social support were particularly satisfactory for the contestants, while the prize structure may not have been optimal. While some teams thought that it was ideal to donate the prize to a charity, others would have liked to receive the prize themselves. The WERO intervention comprised several behaviour change strategies but the design of the study precluded independent assessment of the relative contribution of each. Future controlled studies need to be conducted that allow for assessment of the individual elements, as well as the influence of co-ordinators or coaches. It would be useful to investigate if WERO would be as effective with other populations, for example Indigenous populations in other countries, or pregnant smokers. It is also possible that this intervention could be modified to trigger change in other health related behaviours, such as physical activity and nutrition. Lastly, it would be useful to conduct future work to establish cost-effectiveness and the cost per quitter of running this type of intervention.