Background

Tobacco use remains an urgent and critical issue, causing enormous health and economic burdens.1–3 As the world’s largest producer and consumer of tobacco, China suffers greatly with a high smoking rate (26.6%) and more than 300 million current adult smokers.4 According to the China Adult Tobacco Survey, nearly 50 million smokers intend to quit within 1 year, whereas 90.1% of smokers who have attempted to quit in the past year report not using any cessation aid during the quitting process.4 It suggests that accessible and feasible smoking cessation services are limited, disproportional to the high demand for quitting.

Financial incentives, as one of the behavioural intervention strategies, prompting people to undertake positive behaviour change based on achievable goals.5 Originating from Skinner’s operant conditioning theory, financial rewards operate as positive reinforcers for the expected health behaviour (ie, quitting smoking).6 The Cochrane review published in 2019 has demonstrated that financial incentives significantly improve smoking cessation rates at long-term follow-up,7 despite the fact that most studies included were conducted in developed and high-income countries,8–13 and the review did not include any such study in the Chinese setting.

In China, financial incentives have been applied in various fields of study to reinforce positive health behaviours, including reducing high-risk drug use and injection,14 15 uptaking of test and treatment for infectious diseases,16–18 engagement and retention in addiction-related treatment,19 20 chronic disease self-management21 and improving healthcare service delivery.22 23 To the best of our knowledge, there was only one study in Hong Kong, China investigating the effects of a small financial incentive plus active referral on smoking abstinence among community smokers.24 The study provided some preliminary evidence that financial incentives could be effective for smoking cessation among smokers living within a Chinese context. However, this study took place in Hong Kong, a city that is classified as a high-income economy, in contrast to mainland China, which is classified as a middle-income economy25; another difference is that smoking cessation services in Hong Kong are provided free of charge with good accessibility.26 In addition, it was difficult to disentangle the effects of financial incentives because incentives were combined with other treatment and participants were rewarded for any utilisation of cessation services. Disparities in economic development and healthcare systems, together with the unclear operational mechanisms further complicate the transferability of financial incentive frameworks to smokers in mainland China. Therefore, more evidence is needed to examine whether direct financial rewards for quitting smoking are effective on smoking cessation among smokers living in mainland China.

From the behavioural-economic perspective, financial incentives of similar size may have different effects depending on their design.27 Unlike traditional rewards, a deposit scheme is featured in that it incentivises individuals to earn incremental rewards for desired behaviours, with the accumulated bonus being forfeited if the set target is not achieved.28 29 This strategy is informed by the concept of present bias, a phenomenon in behavioural economics that proposes that individuals are generally more motivated to avoid losses than to seek gains.30 31 Nevertheless, very few studies have compared the effectiveness of deposit and reward programmes, and the results remain inconsistent.32 33 Individuals’ behaviour decisions can differ depending on complex behavioural economic features, such as loss aversion,34 35 delay discounting34 36 and other decision-making preferences,37 warranting further research and discussion.

Previous research on financial incentives has highlighted challenges in maintaining sustained abstinence rates after the monetary stimulation period, potentially due to the temporary nature of external motivators and their impact on intrinsic motivation.38 Therefore, we will incorporate text messages as a component in all study arms, including the control group. Mobile health programmes have proven effective in assisting smoking cessation, particularly in resource-limited settings and low-middle-income countries.39 40 By integrating text messages into all study arms, we aim to leverage their proven ability to influence quitting-related perceptions,41 potentially mitigating the risk of diminishing motivation associated with financial incentives alone. In addition, in spite of the documented effectiveness of financial incentives, smoking cessation rates are usually less than 20%,28 29 suggesting the existence of other impacting factors. Transtheoretical model is featured with processes of change, along with decisional balance (weighing the pros and cons of a behaviour) and self-efficacy.42 While these constructs have been used in smoking cessation interventions,43 their application in the context of financial incentive interventions is limited.

To bridge these gaps in existed literature, we will include constructs attributed to behavioural economics, especially loss aversion to design different incentive schemes and examine their roles across study designs. We will also use the trans-theoretical model and its constructs to interpret how text messages and financial incentives affect smoking cessation.43 44 We will conduct a three-arm cluster randomised controlled study, comparing two financial incentive interventions with the smoking cessation text messages developed by the NCI (National Cancer Institute) among Chinese smokers from worksites. The objectives of this study are to:

(a) Assess the effectiveness of two financial incentive interventions—rewards and deposits, compared with a non-incentive control group and explore the possible effects of different financial incentive scheme designs on smoking cessation outcomes.

(b) Explore the effects of behavioural economic factors such as loss aversion, delayed discounting, plus decisional balance and self-efficacy on smokers’ behaviour change process and successful abstinence.

Methods and analysisOverall study design

This study will be a three-arm cluster randomised controlled trial (figure 1), and the worksite will be used as the cluster. Following the baseline assessment, worksites will be randomly assigned to one of the three arms. Smokers working in the assigned worksite will be classified to the corresponding arm. The two financial interventions and the control conditions will be implemented at the individual level. Follow-up visits are conducted at 1, 3, 6 months post-enrolment. The recruitment and follow-up period of this trial will between April 2023 and February 2024.

Figure 1

Study flow diagram. CO, carbon monoxide.

Inclusion and exclusion criteria

Considering the cluster randomisation design, the first stage of invitation and recruitment takes place at the cluster (worksite) level. Inclusion criteria for the worksite include: (1) consent to enrolled employees participating in questionnaires collecting and exhaled carbon monoxide (CO) testing during working hours; (2) at least 30 employees who are current smokers and (3i) willing to cooperate in the whole process of the trial implementation.

All these criteria will be confirmed during a prerecruitment meeting with a company officer. Following enrolment of the worksite, employees of the selected worksites will be evaluated according to the participants’ criteria. Inclusion criteria for individual smokers include: (a) current smokers; (b) smoking at least 5 cigarettes per week and having ever smoked at least 100 cigarettes in lifetime; (c) exhaled CO level ≥6 ppm; (d) aged 18–65 years old; (e) being a full-time employee at the selected worksite with no plans to leave within the next 6 months and (f) willing to participate in this study and able to provide informed consent.

Participants will be excluded if they meet any of the following exclusion: (a) severe liver and lung dysfunction; (b) received other smoking cessation treatment within 30 days; (c) pregnant or lactating women, or planning for pregnancy within the next 6 months; (d) sever heart, lung, brain or blood system diseases and (e) mental illness and drug addiction. Participants’ eligibility will be verified by trained research staff prior to formal enrolment.

Study setting and recruitment

We will target enterprises in Inner Mongolia, Jiangsu and Hebei, three provinces located in west, east and north China, respectively. Companies and enterprises will be invited to take part in our trial through local bureaus for industry and commerce. We expect to recruit worksites primarily in the mining, manufacturing or energy industries, which are characterised by high labour intensity and high smoking rates. Then an informal conference with administrators from worksites will be held by study personnel, introducing cluster inclusion criteria. Selected worksites are requested to assist with recruiting employees by distributing recruitment materials such as advocacy leaflets, putting up posters and disseminating recruitment notice via internal working networks. The recruiting time is between June 2023 and July 2023.

Randomisation and masking

Selected worksites and participants within them will be divided into three arms in a 1:1:1 ratio using a block-randomisation method. Participants will be randomly assigned with stratification according to study sites and then the time of enrolment will be used as a block factor in each stratum. SAS V.9.4 software package will be used to generate random tables by a biostatistician who will not be involved in the data management and statistical analysis. Due to the nature of our study design, participants in the two intervention groups could not be masked, but they will only be informed of their own financial incentive programmes via WeChat applet (a nested mini programme or application) and not aware of the procedures of the other. The exhaled CO assessor will be masked to treatment allocation throughout the trial. The researchers for money distribution will be unblinded after the completion of the first follow-up visit.

Interventions

In total, there are three arms or randomisation groups, including one control group and two intervention groups. After the assignment, participants in the control group will receive one text message welcoming their participation and reminding them of reading the following text messages. At the same day, participants in two intervention groups will also receive one text message similar to the control group, plus clearly explaining the incentive programmes corresponding to the procedures in each arm.

Control group

Participants will receive one text message per day from the text message library developed by the US NCI over a 3-month intervention period. The text message library was established based on cognitive–behavioural approaches, containing information on adverse impacts of smoking, and encouragement and advice to help quitting smoking. After being translated into Chinese, the text messages have shown promise for assisting smoking cessation among Chinese smokers.45 Text messages will be sent via an applet in WeChat, which is the most popular social media in China.

Reward group

Participants in this arm will receive text messages same as the control group, plus each participant will be rewarded with CNY200 and CNY400 immediately after biochemically verified smoking abstinence at 1-month and 3-month post-enrolment. If the participant continues to smoke at 3-month follow-up visit, he or she will not receive rewards regardless of abstinence status at 1-month follow-up visit. Nevertheless, in cases where participants fail to quit smoking at the 1-month follow-up visit, they will still receive a reward of CNY200 for demonstrating verified abstinence over a period ranging from one to three months. CNY100 equals approximately US$15, which accounts for about 3.3% of monthly disposable income per capita of Inner Mongolia where the trial is conducted.

Deposit group

Participants in this arm will receive text messages same as the control group, plus each participant will be informed that he or she will accumulate CNY200 and CNY400 in the deposit accounts after biochemically verified smoking abstinence at 1, 3 months post-enrolment. The accounts will be reset to zero if the participant continues to smoke or fails to maintain abstinence at 3-month follow-up visit. The participants cannot withdraw or spend the money in the deposit accounts regardless of abstinence status until 3 months post-enrolment. At the 3-month follow-up visit, each participant will receive all the deposits in the account at once after assessments; if one’s account is reset to zero, that is, the participant continues to smoke or fails to maintain abstinence, he or she will not be paid. Similarly, if participants are unable to quit smoking during the 1-month follow-up visit, they will still receive a financial reward of CNY200 for verified abstinence over a duration ranging from 1 to 3 months.

Study proceduresInformed consent procedure

The study will obtain informed consent from participants prior to enrolment and baseline assessment. Before participants’ signing and dating the form, all inquiries will be explicitly and precisely answered by study personnel conveyed in language understandable. In addition to the writing informed consent requested prior to enrolment, key information on voluntary participation with right of refusal, possible risks and confidentiality of records are provided along with the questionnaires each follow-up visit.

Baseline assessment

During the baseline visit, a questionnaire survey and test of exhaled CO are conducted. Questionnaires assess the subject’s basic information, tobacco use conditions and cessation history, nicotine dependence, behavioural economics variables (loss aversion and delayed discounting) and cognitive factors related to smoking cessation (decisional balance and self-efficacy). Participants are also required to be registered for the WeChat applet so they can receive the assigned programme and the following text messages. Seven days after the baseline assessment will be the quit date, from when complete smoking abstinence is urged.

Follow-up assessments

Follow-up visits will be arranged at 1, 3 and 6 months after the enrolment, respectively. Questionnaires collect tobacco use conditions, nicotine withdrawal symptoms, behavioural economics variables and cognitive factors related to smoking cessation. Self-reported smoking abstinence and exhaled CO levels are recorded each time. In the reward group, rewards will be given to abstinent participants in cash immediately when biochemically verified. In the deposit group, participants will be informed of the amounts of deposits accruing to the account at 1-month and 3-month follow-up visit and the total deposits will be given at the end of the 3-month assessment. Follow-up visits will be conducted by trained study personnel each time in a neat and separate office provided by the worksite, where privacy and convenience of participants can be ensured. See table 1 for the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) flow diagram of the schedule of enrolment, intervention and the full list of assessments.

Table 1

SPIRIT flow diagram of enrolment, intervention and assessments

Outcome measuresPrimary outcome

The primary outcome is smoking abstinence rate at 6 months post-enrolment. The outcome is measured by self-reported and biochemically verified through exhaled CO levels (<6 ppm). Exhaled CO levels are examined using a CO monitor (Bedfont Smokerlyzer M-II, Bedfont Scientific, UK). Exhaled CO testing is a simple, non-invasive and immediate method to evaluate smoking status. Previous studies have reported that 6 ppm is the ideal cut-off point to distinguish between smokers and non-smokers.46–48 Sandberg et al proposed 12 ppm as the cut-off value to discriminate recent smokers from short-term abstinent smokers (at least 8 hours).49 It is an absolute necessity to underscore the sensitivity of the test for the purpose of verifying sustained abstinence. Therefore, we adopt a cut-off point of 6 ppm for classification between smokers and non-smokers to confirm smoking abstinence biochemically.

Secondary outcomes

The secondary outcomes include biochemically verified self-reported smoking abstinence at 1 and 3 months. Apart from quitting outcomes, exhaled CO concentration, nicotine withdrawal symptoms and cognitive factors related to smoking cessation (decisional balance, self-efficacy) will also be compared. In addition, the financial costs of each successful abstinence will be assessed. Decisional balance and self-efficacy are measured using the 6-item and 3-item 5-point Likert scales, respectively. Each item was rated from strongly disagree to strongly agree. Examples of pros items include ‘My health will improve after quitting smoking’, while cons items include ‘My social life will be affected after quitting smoking’. An example of a self-efficacy item is ‘I am confident that I can successfully quit smoking’. The scale’s reliability and validity have been verified among adult smokers in China.50

There will not be an explicit restriction on participants’ utilisation of additional quitting methods apart from interventions in this trial, such as pharmacological approaches, for ethical considerations. To disentangle the effect of financial incentives, we will add questions at the follow-up inquiring if the participants have used any other smoking cessation methods and fill in the details. Sensitivity analyses will be conducted for this subset of participants in subsequent analyses.

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

Sample size

We estimate that we will be able to detect abstinence rates of 0.7% in the control group vs 12.7% in either of the incentive-based arms on the basis of previous findings.9 29 The sample size per arm is calculated based on the multiple-group overall difference comparison. Taking type-I error α value of 0.05, the sample size of each group required is 81 to obtain 90% power (β=0.10).We assume the cluster size of 25 for feasibility, and the intraclass correlation coefficient (ICC) is set at 0.022, by taking the average of items related to tobacco use and smoking cessation behaviours (percentage daily smokers; cigarettes/d; Fagerström score; percentage intend to quit in 6 months; percentage intend to quit in 30 days) according to Martinson et al.51 The design effect is 1+(m−1)ICC=1.528, and the sample size is increased by the dropout rate of 20%. Therefore, we need 462 participants, with at least 154 individuals in each arm.

Data management

Double data entry will be performed for the results of exhaled CO testing. Questionnaires will be administrated via online platform and related data will be collected electronically. The principal investigator (PI) of the study (CC) has access to all interim results and authorisation of interim analyses. The PI will make the final decision to terminate the trial. On completion of the follow-up visits and data collection, all data files will be stored securely in password-protected computer systems to prevent destruction and be managed by a dedicated person. A data monitoring committee is established to monitor the research progress and data quality, entirely independent from the sponsor. Before data analyses, personal information will be anonymised. To protect confidentiality, no individual case analyses will be performed.

Statistical analysis

Descriptive analyses will be performed for baseline characteristics, with mean, SD for normally distributed continuous variables and median, IQR for non-normally distributed ones and proportions for categorical variables. χ2 or t-test analyses will be selected on applicable conditions to compare the distributions of baseline data. All control variables will be compared across arms to weigh up intergroup balance, including sociodemographic characteristics and the other variables of interest or known to be associated with the outcomes.

We will apply the generalised linear mixed-effects models to evaluate the effectiveness of the intervention, using the intention-to-treat approach. The cluster will be considered as a random effect to adjust for attributes of each individual cluster as a function of covariate.

Several sensitivity analyses will be applied to assess the robustness of the intervention effects estimated. First, a per-protocol analysis will be conducted instead of the initial intention-to-treat approach, in which participants ceasing to follow the assigned treatment will be censored.52 Second, we include participants who will be censored when they report unanticipated events causing abstinence such as diseases and pregnancy into the full analysis samples. Third, we will deal with missing data using an alternative approach. Participants failing to show up will present missing observations in the dataset, and we will consider them as non-abstinent in the sensitivity analysis. Fourth, we will use a substitute definition of sustained abstinence. We will treat participants reporting e-cigarettes use as long as no cigarettes used as smoking cessation.

We will compare point-prevalence abstinence rate and only self-reported sustained abstinence rate using similar mixed-effects models. The factors influencing the effectiveness of the intervention will be assessed by multifactor logistic regression models. The structural equation models will be constructed to assess the relationships and path coefficients between the factors mentioned above and smoking cessation behaviours.

Discussion

This study is designed to evaluate the effectiveness and feasibility of two financial incentives on smoking abstinence compared with a no-incentive group. The benchmark is the delivery of smoking cessation text messages during 3 months, which has proven effective in previous research. The impacts of two financial interventions are expected to differ based on behavioural-economic theories and specific monetary values in Chinese socio-cultural contexts. This is the first study, of which we are aware, testing financial incentives on quitting smoking targeting at smoking employees in workplaces in China.

However, there are several potential limitations in this trial. The value of exhaled CO in this study can be deceptive due to the half-time of all biomarkers and relatively longer follow-up period, that is, the value may be lower than the cut-point if a smoker has refrained from smoking for a short term, for example, 8 hours.49 Nevertheless, we consider testing exhaled CO to verify self-reported abstinence as feasible for its widely use and the sensitive cut-point adopted, and participants will not be informed of specific half-time. Also, it remains beyond the current design to assess whether smoking cessation can be maintained and consolidated after 6 months (the end of our follow-up) due to budgetary constraints.

This study will advance the literature to propose financial incentive-motivated smoking cessation programmes in high-demand low-resource countries. We anticipate the findings can provide implications for practice considering the large number of current smokers in China, specifically spurring both private sector and large enterprises to consolidate incentive-based approaches into health welfare packages to reduce productivity loss due to smoking. Moreover, the incentive projects are expected to be portable to other settings and even a wider range of national tobacco control strategies if applicable.

Ethics and dissemination

This trial protocol has been approved by the Ethics Committee of Peking University Health Science Centre (date: 23 February 2023; ethical approval number: IRB00001052-22172; protocol version 2.0). Written informed consent will be obtained from participants prior to enrolment. We will communicate with investigators, trial participants, trial registries and journals and submit related reports following the rules of the ethics committee for any important protocol modification during the trial process. All procedures in this study will be carried out in accordance with the ethics committee guidelines and regulations and the study protocol will be conducted in accordance with the Declaration of Helsinki. Results and findings of this trial will be disseminated in peer-reviewed journals and professional conferences.