Brief Advice, Nicotine Replacement Therapy Sampling, and Active Referral for Expectant Fathers Who Smoke Cigarettes: A Randomized Clinical Trial

Key Points

Question  Is brief advice combined with a 1-week nicotine replacement therapy (NRT) sample and active referral an effective intervention for smoking cessation among expectant fathers who smoked cigarettes?

Findings  In this randomized clinical trial of 1053 participants recruited from prenatal clinics in Hong Kong, providing brief advice with offers of a 1-week NRT sample and active referral compared with brief advice alone was associated with higher biochemically validated tobacco abstinence at 6 months in the intervention group vs the group of brief advice alone (6.8% vs 3.6%).

Meaning  Results of this trial showed that a proactive, combined intervention nearly doubled the odds of validated quitting among expectant fathers who smoked.

Importance  Pregnancy presents an opportunity to engage expectant fathers in smoking cessation, but evidence from randomized clinical trials is scarce.

Objective  To evaluate the effectiveness of a proactive, combined intervention for smoking cessation in expectant fathers.

Design, Setting, and Participants  This pragmatic randomized clinical trial in prenatal clinics in 7 public hospitals in Hong Kong proactively recruited and enrolled 1053 participants from October 10, 2018, to February 8, 2020. Included male adults were 18 years or older, smoked cigarettes daily in the past 3 months, had partners who were pregnant and nonsmoking in the past 30 days, and had a landline or mobile telephone number for follow-up. These participants were randomized to either the intervention group or the control group. The primary analyses used an intention-to-treat approach.

Interventions  The intervention group received brief cessation advice, a 1-week free sample of nicotine replacement therapy (NRT), and active referral to a community-based smoking cessation service. The control group received only brief cessation advice with a leaflet on the hazards of perinatal exposure to tobacco smoke and the toll-free quitline telephone number.

Main Outcomes and Measures  The primary outcome was biochemically validated tobacco abstinence at 6 months after intervention initiation defined by an exhaled carbon monoxide level of 3 parts per million or lower. The secondary outcomes included self-reported 24-week continuous abstinence at 6 months after intervention initiation as well as 7-day point prevalence abstinence, use of any NRT, and use of a smoking cessation service at 3 and 6 months after intervention initiation.

Results  All 1053 randomized participants were male adults with a mean (SD) age of 33.8 (6.9) years. The retention rate at 6-month follow-up was 80.7%. The primary outcome of biochemically validated tobacco abstinence at 6 months after intervention initiation was significantly higher in the intervention group than in the control group (6.8% [36 of 527 participants] vs 3.6% [19 of 526]; odds ratio [OR], 1.96; 95% CI, 1.11-3.46; P = .02). The main secondary outcomes of self-reported 24-week continuous abstinence at 6 months (OR, 1.87; 95% CI, 1.08-3.23; P = .03) and 7-day point prevalence abstinence at 3 months (OR, 1.48; 95% CI, 1.05-2.09; P = .03) and 6 months (OR, 1.74; 95% CI, 1.29-2.34; P < .001) were also significantly higher in the intervention group. The intervention group had a significantly higher increase in perceived family harmony (score range, 0-10, with a higher score indicating a higher level of harmony) from baseline to 6 months (B = 0.28; 95% CI, 0.063-0.50; P = .01).

Conclusions and Relevance  This trial found that combining brief advice with a 1-week sample of NRT and referral to smoking cessation programs nearly doubled the odds that expectant fathers who smoked would achieve validated abstinence compared with providing brief advice alone. The intervention was also effective in promoting family harmony.

Trial Registration  ClinicalTrials.gov Identifier: NCT03671707

Secondhand smoke exposure in nonsmoking pregnant women is associated with adverse fetal and child outcomes, such as stillbirth, birth defects, and neurodevelopmental delay.^1-3 Many pregnant women are exposed to secondhand smoke in high-income (eg, 35% in the US)⁴ and low- to middle-income countries (29%),⁵ including China (>39%).⁶ Pregnancy presents an opportunity to promote quitting among expectant fathers who smoke cigarettes,⁷ but a recent systematic review⁸ found only 1 randomized trial of a smoking cessation intervention directed at expectant fathers with nonsmoking pregnant partners. The World Health Organization has called for more research on engaging expectant fathers in smoking cessation.³

Delivering the 5 As (Ask, Advise, Assess, Assist, and Arrange follow-up) intervention for smoking cessation has remained a challenge in busy clinical settings, in which less than half of the smokers in primary care were offered assistance to quit and followed up.^9,10 An approach comprising brief cessation advice and referral to an external smoking cessation service was found effective in increasing treatment uptake and abstinence.^11-13 Offering 1 to 2 weeks of free nicotine replacement therapy (NRT) samples can promote cessation^14,15 primarily by enhancing a person’s motivation and self-efficacy to quit.¹⁶ These interventions are short, low cost, and simple to deliver and thus can potentially support smoking abstinence in expectant fathers who accompany their partners for prenatal visits. A PubMed search that we performed using the Medical Subject Headings keywords smoking and pregnancy did not identify any trial that tested similar strategies in expectant fathers.

In Hong Kong, the most westernized city in China, nearly 30% of mothers with a newborn reported living with a partner who smoked cigarettes.¹⁷ As in many other places, expectant fathers in Hong Kong typically do not receive any smoking cessation advice or support during prenatal visits. In the current study, we evaluated the effectiveness of a proactive, combined intervention for smoking cessation called brief advice, nicotine replacement therapy sampling, and active referral (BANSAR), in expectant fathers.

We conducted a pragmatic randomized clinical trial in the prenatal clinics of 7 public hospitals (Kwong Wah Hospital, Queen Elizabeth Hospital, Queen Mary Hospital, United Christian Hospital, Tuen Mun Hospital, Pamela Youde Nethersole Eastern Hospital, and Princess Margaret Hospital), whose total catchment area is about 82% of the Hong Kong population.¹⁸ The trial was fairly pragmatic according to the Pragmatic Explanatory Continuum Indicator Summary 2 (PRECIS-2) tool,¹⁹ and the trial protocol (Supplement 1) was published elsewhere.²⁰ The trial was approved by the institutional review board at each hospital site. All participants provided written informed consent. We followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

From October 10, 2018, to February 8, 2020, research nurses or assistants proactively approached and screened for eligibility the expectant fathers at the prenatal clinics. Pregnant women whose partners were cigarette smokers but not present were encouraged to bring their partners to the next prenatal visits and given a flyer with a description of the study, contact information of the research team, and a Quick Response code that was linked to an online registration form. To be eligible, both the expectant fathers and mothers had to be Hong Kong residents, 18 years or older, living together in the past 7 days, and able to communicate in Cantonese or Mandarin. Included male adults smoked cigarettes daily in the past 3 months, had partners who were pregnant and nonsmoking in the past 30 days, and had a landline or mobile telephone number for follow-up. We excluded smokers with a history of angina, arrhythmia, or acute myocardial infarction because they may need to use NRT with caution. Those with psychiatric diseases, who were taking regular psychotropic medications, or who used a smoking cessation aid or program in the past 3 months were also excluded.

Recruitment was projected to end in March 2020 but was suspended on February 10, 2020, and then terminated on October 21, 2020, because of the local COVID-19 outbreak (eMethods in Supplement 2). Thus, the final sample size was slightly smaller than planned.

A biostatistician who was not a member of the research team generated the randomization list with a 1:1 allocation ratio and random permuted block sizes of 2, 4, or 6 to achieve similar numbers of participants in both intervention and control groups (Figure). The randomization was concealed by using sequentially numbered, opaque sealed envelopes, which were prepared by 1 of us (T.T.L.), who was not involved in participant recruitment. After obtaining written informed consent and collecting baseline data, we opened a sequentially numbered, opaque sealed envelope to determine the group randomization. Blinding of the participants and the research nurses and assistants was not possible because of the nature of the behavioral interventions. Statistical analyses were performed by 2 of us (T.T.L. and Y.S.W.), who were not involved in data entry and were blinded to the group randomization until all prespecified analyses were completed.

The pregnant women did not receive any intervention except advice on the health risks of perinatal passive smoking. Expectant fathers who were randomized to the intervention group received BANSAR according to the AWARD (Ask, Warn, Advise, Refer, and Do it again) model,¹³ which was adapted from the 5 As intervention for smoking cessation. Research nurses or assistants asked expectant fathers at the clinics about their smoking behaviors (Ask). Current cigarette smokers were then invited to test their exhaled carbon monoxide level using the Smokerlyzer (Bedfont Scientific Ltd). The readings were shown to the smokers as a warning about the health risks of secondhand smoke exposure for pregnant women, the fetus, and young children (Warn). The smokers were then advised to quit smoking as soon as possible and to enroll in the study (Advise).

Participants in the intervention group were offered a 1-week sample of NRT patch or gum in its original packaging.²¹ A previous trial showed no difference in quit rate between 1 week and 2 weeks of NRT sampling.²² Participants without a preference on the NRT form and who smoked fewer than 10 cigarettes per day were given 2-mg gum; those who smoked 10 to 20 cigarettes per day received 14-mg patches, and those who smoked 20 or more cigarettes per day received 21-mg patches. Research nurses or assistants provided an instruction card that described how to use the NRT patch or gum and encouraged the participants to make quit attempts without the pressure of quitting successfully.^14,15 In addition, participants were offered a referral to a community-based smoking cessation service of their choice (Refer).¹³ These services are free to Hong Kong residents and provide evidence-based cessation treatment, including counseling and full-course pharmacotherapy as appropriate.¹³ Participants were given a leaflet about available services and were encouraged to select a service. Telephone numbers of those who agreed to be referred were sent to the practitioner at the selected service, who subsequently contacted the participants for further treatment. Participants also received 2 telephone boosters (lasting about 2 minutes each) from a research nurse within the first month after baseline (enrollment) to address any issue related to the NRT (eg, adverse effects) and the smoking cessation service (Do it again).

Participants who were randomized to the control group received only brief cessation advice (Ask, Warn, Advise) with a standard leaflet on the hazards of perinatal exposure to tobacco smoke and the toll-free quitline telephone number in Hong Kong. Through the quitline, participants in the control group could access the same smoking cessation services to which participants in the intervention group were actively referred.

Data were collected in person at baseline and through telephone interviews at 3 and 6 months after randomization (intervention initiation). The primary outcome was biochemically validated tobacco abstinence at 6 months after intervention initiation,²³ defined by an exhaled carbon monoxide level of 3 parts per million or lower, as measured with the Smokerlyzer.²⁴ Participants who self-reported having quit tobacco use for 7 days or longer at the 6-month follow-up were invited for the validation test (eMethods in Supplement 2). Those who agreed to the validation test were given HK $300 (approximately US $38) in cash for their time and traveling expense.

Secondary outcomes included self-reported 24-week continuous abstinence at 6 months after intervention initiation as well as 7-day point prevalence abstinence, use of any NRT, and use of a smoking cessation service at 3 and 6 months after intervention initiation. To ensure consistent outcome measurement in both groups, we did not differentiate the outcome from using the NRT sample that we provided and NRT from other sources. Other outcomes regarding those who continued to smoke included self-reported smoking reduction, defined by at least a 50% decrease from the baseline cigarette consumption at the 3- and 6-month follow-up, and changes in cigarette dependence and readiness to quit from baseline to the 3- and 6-month follow-up.

To assess the potential family conflict induced by the intervention,^3,7 we asked participants to rate the following question on a scale of 0 (not at all) to 10 (very much) at baseline and 6-month follow-up: “To what extent do you think your family is harmonious/happy?”²⁵ Family harmony and happiness are validated indicators of family relationship quality among Hong Kong Chinese.^26,27

For process evaluations, anonymized data on smoking behaviors were collected from eligible smokers who refused to participate in the study. At the 3-month follow-up, we asked participants to rate the appropriateness and helpfulness of the brief cessation advice and leaflet on a scale of 1 (not appropriate/helpful at all) to 5 (very appropriate/ helpful). At the 6-month follow-up, we assessed intervention satisfaction by asking participants to rate the following question on a scale of 0 (not at all) to 10 (very much): “To what extent would you recommend the treatment program to other expectant fathers?”

The sample size was estimated by using evidence from a previous trial of brief advice and active referral,¹³ which found a 6-month biochemically validated abstinence rate of 5.0% in the control group and a treatment effect (odds ratio [OR]) of 1.85 by intention-to-treat analysis. With a randomization ratio of 1:1 and power of 0.80, we determined that 574 participants per group (1148 in total) were needed to detect a significant treatment effect at a 2-sided 5% level of significance.

All analyses were performed according to a prespecified statistical analysis plan²⁰ using Stata/MP, version 15.1 (StataCorp LLC). A 2-sided P < .05 indicated statistical significance. The primary analyses used an intention-to-treat approach, and participants with missing outcomes were assumed to have had no change in smoking behaviors from baseline. We used logistic regression to compare the primary and secondary outcomes between the intervention and control groups. We used descriptive statistics to compare other outcomes in participants who continued smoking and participant ratings of the brief advice, leaflet, and intervention between the 2 groups.

Four planned sensitivity analyses were conducted for the abstinence outcomes: (1) generalized estimating equation modeling to adjust for potential clustering by hospital sites; (2) multivariable logistic regressions to adjust for potential imbalances in baseline characteristics; (3) multiple imputations by chained equations to impute missing abstinence data, with inferences drawn from 50 imputed data sets; and (4) complete case analyses (eMethods in Supplement 2).

To evaluate intervention consistency, we assessed the treatment effect on the primary outcome in prespecified subgroups of cigarette dependence: readiness to quit within 30 days, any previous quit attempt, the stage of pregnancy (in trimester) during recruitment, living with another smoker, and ever smoking status of the expectant mother. We calculated the P value for multiplicative interaction by including an interaction term in the logistic regression models. We used linear mixed-effects regressions with time multiplied by group interactions to estimate the differences in change in perceived family harmony and happiness from baseline to 6 months between the 2 groups, unadjusted for covariates per the primary analyses (eMethods in Supplement 2).

We screened for eligibility 1904 expectant fathers who smoked. Of the 1415 smokers identified as eligible, 1053 (74.4%) consented to participate and were randomized to either the intervention group (n = 527) or control group (n = 526) (Figure). All 1053 randomized participants were male adults with a mean (SD) age of 33.8 (6.9) years. Those who refused participation had similar smoking characteristics (eTable 1 in Supplement 2).

The median (interquartile range [IQR]) time spent on recruiting a participant, collecting baseline data, and delivering baseline interventions was approximately 12 (9-15) minutes. Most participants ranged in age from 26 to 45 years (903 [86.3%]), had moderate to high heaviness of smoking (328 [31.1%]), had never tried to quit (404 [38.4%]), and were not ready to quit in 30 days (840 [79.8%]). The baseline characteristics were similar between the 2 groups (Table 1).

The retention rates at 6 months after intervention initiation did not differ between the 2 groups (81.6% vs 79.8%; P = .47). Younger age was associated with attrition at the 6-month follow-up (eTable 2 in Supplement 2). The rate of participation in biochemical validation was similar between the 2 groups (24.9%; P = .45), and 1 participant from each group had a failed validation. By intention-to-treat analysis, the primary outcome of biochemically validated abstinence at the 6-month follow-up was 6.8% (36 of 527 participants) in the intervention group vs 3.6% (19 of 526) in the control group (OR, 1.96; 95% CI, 1.11-3.46; P = .02) (Table 2). The secondary outcomes of self-reported 24-week continuous abstinence at 6-month follow-up (OR, 1.87; 95% CI, 1.08-3.23; P = .03), 7-day point prevalence abstinence at 3-month (OR, 1.48; 95% CI, 1.05-2.09; P = .03) and 6-month (OR, 1.74; 95% CI, 1.29-2.34; P < .001) follow-up, 24-hour quit attempt at 3-month (OR, 1.41; 95% CI, 1.08-1.80; P = .008) and 6-month (OR, 1.52; 95% CI, 1.19-1.94; P < .001) follow-up, and use of any NRT at 3-month (OR, 22.6; 95% CI, 11.4-45.0; P < .001) and 6-month (OR, 27.7; 95% CI, 14.4-53.1; P < .001) follow-up were also significantly higher in the intervention vs control group (Table 2).

The use rate was low overall (2.8% [n = 15] at 3-month follow-up and 4.7% (n = 25) at 6-month follow-up in the intervention group) but was not significantly higher in the intervention than in the control group at 3-month (OR, 2.17; 95% CI, 0.88-5.37; P = .09) and 6-month (OR, 1.70; 95% CI, 0.88-3.26; P = .11) follow-up.

Among expectant fathers in the intervention and control groups who continued to smoke, 200 (24.3%) reported reducing their cigarette consumption by half or more at the 6-month follow-up (Table 3). Cigarette dependence and readiness to quit slightly decreased in both groups, but the decrease in cigarette dependence at the 6-month follow-up was significantly greater in the intervention than in the control group (−0.37 vs −0.15; P = .003).

The differences in abstinence outcomes based on the generalized estimating equation models, multivariable models, multiply imputed data, and complete case analyses were all significant, with effect sizes that were similar to those of the primary analyses (eTables 3 and 4 in Supplement 2). The clustering effect of hospitals was negligible (intracluster correlation coefficients, <.001).

Favorable treatment effects were observed in subgroups of all prespecified baseline characteristics without significant interaction effects. For example, the intervention effects on validated abstinence were similar between participants who were ready to quit within 30 days vs after 30 days (OR, 2.01 [95% CI, 0.80-5.00] vs 2.01 [95% CI, 0.96-4.21]; P for interaction >.99) (eTable 5 in Supplement 2). The mean scores of perceived family harmony and happiness in both groups slightly increased from baseline to 6 months after intervention initiation. The increase in perceived family harmony was significantly higher in the intervention than in the control group (B = 0.28; 95% CI, 0.063-0.50; P = .01) (eTable 6 in Supplement 2).

The perceived appropriateness (mean [SD] score, 4.2 [0.8] vs 4.1 [0.9]; P = .11) and helpfulness (mean [SD] score, 3.4 [1.0] vs 3.4 [1.0]; P = .97) of brief cessation advice were similar between the intervention and control groups, as was the perceived helpfulness of the self-help leaflet among those who read it (mean [SD] score, 2.6 [1.1] vs 2.7 [1.1]; P = .72) (Table 4). Intervention satisfaction was significantly higher in the intervention than the in control group (mean [SD] score, 6.3 [2.2] vs 5.7 [2.6]; P = .004).

This pragmatic trial showed that BANSAR, a simple intervention that combines brief cessation advice with offers of a 1-week NRT sample and referral to a smoking cessation service, was effective compared with brief advice alone in increasing 6-month biochemically validated abstinence in expectant fathers who smoked. Its effect size was similar to that of the self-reported 24-week prolonged abstinence and 7-day point prevalence abstinence. The intervention group also showed higher intervention satisfaction compared with the control group. The real-world effect of BANSAR would likely be greater because expectant fathers do not receive any cessation advice in usual practice in Hong Kong, whereas the control group in this study at least received brief advice.

A previous trial that targeted expectant fathers did not find intensive couple-based counseling effective in increasing validated abstinence.²⁸ The current trial showed the effectiveness of a combined intervention that can be readily implemented in prenatal clinics. We could not disentangle the individual effect of NRT sampling from the effect of active referral; both of these approaches appeared to increase abstinence by promoting the use of evidence-based treatment. The NRT samples might serve as a taste of cessation treatment and abstinence for smokers who have not made a commitment to quit, which subsequently may prompt some smokers to initiate cessation. This hypothesis was corroborated by the higher rates of 24-hour quit attempt and NRT use in the intervention than in the control group. The results of this trial were consistent with those of a recent trial conducted in primary care clinics (despite the differences in study designs), which found that NRT sampling was effective in increasing self-reported abstinence at 6 months after treatment initiation (OR, 1.5).¹⁵ The greater effect size in the current trial (OR, 1.96) might be attributed to the addition of active referral in the intervention. Smoking cessation service use was low (2.8%-4.7%) but meaningfully higher in the intervention than in the control group (OR ranged from 1.70 to 2.17). Given that smoking cessation service use was associated with quitting,^29,30 active referral likely contributed to part of the treatment effect by increasing service use in the intervention group.

Communicating the health risk of perinatal secondhand smoking can fuel conflicts between pregnant women and expectant fathers.^3,7 Concern about harming family relationships is a known barrier to nonsmoking women encouraging the male smokers in the family to quit their cigarette use.³¹ In contrast, we found that family relationships improved in both intervention and control groups at 6 months after intervention initiation, with a small yet significant effect on perceived family harmony. A previous study showed that former smokers had a better perception of family well-being than current smokers.²⁷ The higher quit rate in the intervention vs the control group in this trial might explain the effect on family harmony.

We used a proactive approach to recruit and enroll smokers with broader characteristics, rather than a reactive approach (eg, advertisement). Hence, we included many smokers who were not ready to quit smoking in 30 days (79.8% of participants) and hence were unlikely to seek help with quitting. The smoking characteristics were similar between participants and individuals who refused to participate, which suggested that the cohort might be representative of the target population. Favorable treatment effects were observed in participants with different baseline characteristics without significant effect modification, supporting BANSAR as an appropriate intervention for most expectant fathers who smoke cigarettes.

Not engaging expectant fathers in smoking cessation is an important omission in tobacco control. Globally, the smoking prevalence is much higher in men than in women (25.0% vs 5.4%),³² and 35% of nonsmoking women and 40% of children are exposed to secondhand smoke.³³ A previous study did not find an association between paternal smoking avoidance (eg, smoking ≥3 m away from the children) as well as maternal protective actions (eg, encouraging the father not to smoke near the children) and lower levels of salivary cotinine, a biomarker of secondhand smoke exposure, in infants.³⁴ Pregnancy can motivate quitting in expectant fathers to protect pregnant women and children from passive smoking. It also presents an opportunity to help expectant fathers quit smoking when they are still relatively young and healthy, a period in which they are less likely to consult a doctor and to benefit from opportunistic cessation intervention in primary care. Smoking cessation by the age of 40 years could avert the risk of smoking-related death by nearly 90%, compared with continued smoking.³⁵ Providing brief cessation intervention to expectant fathers should be a part of routine practice in prenatal care.

This trial has some limitations. First, the COVID-19 outbreak necessitated the early termination of trial recruitment, which slightly decreased the statistical power from 0.80 to 0.77. Second, bias from incomplete follow-up could not be excluded despite the high retention rate (80.7%). However, the results from multiple imputations were consistent with the results from the primary analyses. Third, the rate of participation in biochemical validation was low but was similar between the intervention and control groups. This finding is commonly reported by real-world trials of brief intervention with minimal contacts between practitioners and participants.^13,29,36 The social pressure to protect pregnant women and children from secondhand smoking could also lead to the overreporting of quitting by some participants. Nevertheless, the effect sizes of validated and self-reported abstinence were similar. Fourth, this trial enrolled expectant fathers who were younger and thus tended to smoke less than the smokers in the general population. The applicability of the study findings to smokers with different characteristics is uncertain.

This trial found that providing brief advice combined with offers of a 1-week sample of NRT and referral to smoking cessation services (BANSAR) nearly doubled the odds of biochemically validated abstinence compared with brief advice alone in proactively recruited expectant fathers who smoked cigarettes. This intervention was also effective in promoting family harmony. Further studies are warranted to evaluate BANSAR in smokers with different characteristics and in other settings.

Accepted for Publication: April 25, 2021.

Published Online: June 14, 2021. doi:10.1001/jamainternmed.2021.2757

Corresponding Author: Man Ping Wang, PhD, School of Nursing, The University of Hong Kong, Hong Kong Special Administrative Region, China (mpwang@hku.hk).

Author Contributions: Drs Luk and Wang (principal investigator) had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Luk, Lam, Kwa, Tang, Lee, Hsieh, Li, Wang.

Acquisition, analysis, or interpretation of data: Luk, Lam, W. C. Leung, K. Y. Leung, Cheung, Kwa, Siong, Hsieh, Wu, Wang.

Drafting of the manuscript: Luk, Lam, Kwa, Tang, Lee, Wang.

Critical revision of the manuscript for important intellectual content: Luk, Lam, W. C. Leung, K. Y. Leung, Cheung, Kwa, Siong, Hsieh, Wu, Li, Wang.

Statistical analysis: Luk, Kwa, Wu, Wang.

Obtained funding: Luk, Lam, Kwa, Li, Wang.

Administrative, technical, or material support: Luk, W. C. Leung, K. Y. Leung, Cheung, Kwa, Siong, Tang, Hsieh, Wang.

Supervision: W. C. Leung, Cheung, Kwa, Tang, Wang.

Conflict of Interest Disclosures: None reported.

Funding/Support: This research was funded by grant 15162691 from the Health and Medical Research Fund of the Food and Health Bureau, The Hong Kong Special Administrative Region Government.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 3.

Additional Contributions: We thank the participants for their contributions; the clinicians for their assistance; and the Smoking Cessation Research Team, School of Nursing, The University of Hong Kong, for implementing the trial.