Effectiveness of Ecological Momentary Assessment Based Smoking Cessation Intervention

Study design:

The proposed study comprises a 2-arm (allocation ratio 1:1) RCT to assess the efficacy of the EMA-based smoking cessation intervention and the EMA-based observational study. Throughout the 7-day participation, all recruited participants will complete an EoD survey and 4 time-based system-triggered EMA each day. The surveys will be prompted by a smartphone application (app) installed in their own mobile phones. The prompting time will be personalized and pre-scheduled by the participants. The information collected in the app will be used to develop an EMA-based intervention including a quit plan and 10-week quitting reminders and for an EMA-based observational study. I will randomly allocate half of the participants (intervention group) to receive the EMA-based intervention. All participants will then be contacted for a follow-up at 3- and 6-month.

Subjects:

This proposed study aims to recruit daily tobacco users, regardless of intention to quit.The inclusion criteria are: (1) daily consumption of tobacco products (including traditional cigarettes, electronic cigarettes, and heated tobacco products) in the past week; (2) age ≥18 years; (3) owning a mobile smartphone with internet access; (4) staying in Hong Kong during the 1-week EMA study period; and (5) able to read and write Chinese. Participants' smoking status will be confirmed by measuring exhaled carbon monoxide to be 4ppm or above, or saliva cotinine to be 30ng/ml or above. Because the proposed study aims to promote smoking cessation in smokers who do not plan to use smoking cessation aids, we will exclude smokers who (1) plan to use smoking cessation services or medication in the coming month or (2) using smoking cessation services, or using nicotine replacement therapy in the past 7 days. Smokers who reported having mental illnesses and female smokers who are pregnant are also excluded as they need more specific smoking cessation intervention.

Recruitment procedures:

Recruitment will be conducted both offline and online. In the outdoor recruitment, the recruitment staff first approach and distribute souvenirs (e.g., a tissue pack) to the smokers at outdoor smoking hotspots. If the smoker is willing to accept souvenirs and talk to the recruitment staff, the staff will ask further questions related to eligibility. Eligible participants will: (1) be given more detail about this project, (2) be invited to provide written consent, (3) be assisted in installing EMA app on their mobile phone and setting EMA schedule, (4) complete baseline questionnaire by paper form or Qualtrics. Due to the time-limited at outdoor recruitment, participants can complete baseline questionnaire after the recruitment session.

Research assistants will also send mass emails or advertisements through social media (e.g., Facebook, Instagram). Potential participants who are interested in this project can scan the QR code on the advertisement and complete the online application form (using Qualtrics). Our recruitment staff makes appointments with the potential participants to complete the recruitment procedures.

Ecological momentary assessment:

An EMA app will be developed to document all smoking cues and smoking-related behaviors. After installing the EMA app on participants' smartphones, participants need to input the five last digit numbers of their telephone number as an identifier and select the date they would like to start the EMA survey. Participants need to complete 5 time-based EMA surveys per day for 7 consecutive days, each EMA survey with an interval of three hours.

The EMA questions on smoking cues cover 4 domains: emotional, social, pattern, and withdrawal cues. Smoking-related behaviors include (1) smoking tobacco (traditional cigarettes, heated tobacco products, and electronic cigarettes) during the past 3 hours, (2) nicotine craving during the past 3 hours and type of craving (physical, psychological or social craving), (3) purchasing tobacco products during the past 3 hours. Whether the smoking-related behaviors have been due to the exposure to the pro-smoking cues will also be asked. Sleep quality will be asked by (1) did you experience insomnia problem (difficulty falling asleep, difficulty staying asleep, problems waking up too early) for last night (none, mild, moderate, severe, very severe), (2) how satisfied are you with your last night sleep pattern (score range from 0 very satisfied to 4 very dissatisfied) in first EMA for each day. 'Time to consume the first cigarette after waking' will be asked at the second EMA. Tobacco control policy, including health warnings on their tobacco pack; POS tobacco displays; smoking hotspots, and daily tobacco consumption, will be asked at the last EMA for each day.

Intervention:

Based on the EMA of each participant, a trained research nurse will review their (1) nicotine dependence; (2) intensity and frequency of reported cravings; and (3) the impact of exposure to pro-smoking cues on smoking behaviors. A personalized quit plan will be developed based on tailored suggestions from the US clinical practice guideline for smoking cessation, Smoking Cessation Information Kit (published by Department of Health), the theory of Health Action Process Approach, and the self-determination theory. Our recent studies showed that interventions based on these theories are effective. Half of the randomly selected participants (intervention group) will receive this quit plan via a nurse-led phone call, email, and WhatsApp. The nurse will then design a quit plan including (1) self-help techniques on handling craving; (2) suggestions to avoid the reported pro-smoking cues; (3) whether to quit progressively or abruptly; (4) referral for existing smoking cessation service; and (5) instruction of using over-the-counter (OTC) nicotine replacement therapy (NRT), if OTC-NRT is preferred by the participant. Participants will then receive tailored quitting reminders via WhatsApp and voluntarily discuss their quit plan with the nurse in the subsequent 10 weeks.

Randomization, blinding, and allocation concealment:

The research staff will randomly assign participants on an individual level to intervention and control group using a 1:1 allocation ratio by sequentially numbered, opaque sealed envelope method. The principle will generate a list of random numbers using the random function of Excel. Then prepare about 450 identical, opaque, sealed, A5-sized envelopes, with a unique 3-digit serial number on the cover of each envelope as an identifier, and put the allocation code into the envelopes. All recruitment staff and participants will be concealed to the group allocation at recruitment. After a participant consents to the trial, the research staff will retrieve his/her information and open 1 envelope to allocate the recruited participant according to the card inside.

Participants and research staff of intervention delivery will not be blinded to the intervention because the intervention is behaviors. However, assessors of the follow-up and the research investigators will not be involved in the recruitment and intervention delivery and will be blinded to the group allocation (i.e. single-blinded design).

Follow-up All participants will be followed up via telephone by an allocation-blinded interviewer at 3- and 6-month after consenting to participation. Only the participants who report abstinence in the past 7 days will be invited to measure their exhaled CO with a Smokerlyzer and/or saliva cotinine level with a Cotinniene Saliva Test Device at their residence, workplace, or nearby, according to their preference.

Sample size:

To date, there was no previous RCT that used EMA results to guide personalized nurse-led telephone counseling and instant messaging for smoking cessation. The effect size for personalized counseling and instant messaging was derived from a network meta-analysis. The sample size was calculated based on an anticipated intervention effect of an OR of 2.92 from the meta-analysis of tailored text messaging combined with counseling for smoking cessation. One of our previous RCTs showed that the prevalence of 3-month biochemically validated tobacco abstinence among Hong Kong smokers was 4% in the group receiving brief advice. To detect a significant difference of quit rate between two groups with a power of 80% (to reduce type II error) at 5% significant level (type I error), a total of 444 participants (allocation ratio 1:1; 222 vs 222) is needed.

Data analysis:

The primary analysis will examine the effect of EMA-based intervention on biochemically validated abstinence at 3-month and the difference in the IBC-S score between the two groups. Intention-to-treat approach will be used to include all consented participants in the analysis and assume non-respondents at the 3-month follow-up as smokers and no change on the IBC-S score. A generalized estimating equation (GEE) analysis with exchangeable structure for the correlation matrix will be used for biochemically validated abstinence after adjusting unbalance baseline demographics and other variables that have been reported to be associated with smoking cessation. We will use linear regression adjusted for imbalance socio-demographics assessed at baseline to compare the score of IBC-S between intervention and control groups. Multiple imputation will be used to deal with missing values for baseline socio-demographics. All secondary and ancillary outcomes will be analyzed with either GEE (binary outcomes) or linear mixed models (continuous outcomes). Subgroup analysis of comparing the primary and secondary outcomes between intervention group participants who received both nurse-led phone calls plus instant messaging and instant messaging only will be conducted.