Online Social Support for the Prevention of Smoking Relapse: A Content Analysis of the WhatsApp and Facebook Social Groups

Abstract

Background:

Online social groups have been increasingly used for smoking cessation intervention.

Introduction:

This study aimed to explore the social support components of the online discussion through WhatsApp and Facebook, how these components addressed the need of relapse prevention, and how the participants evaluated this intervention.

Materials and Methods:

We coded and analyzed the posts (N = 467) by the 82 recent quitters in WhatsApp and Facebook social groups, who were recruited from the eight smoking cessation clinics in Hong Kong to participate in a pragmatic randomized trial of relapse prevention. Participants' postintervention feedback was collected from the 13 qualitative interviews after the intervention.

Results:

The WhatsApp social groups had more participants' posts than the Facebook counterparts. The participants' posts in the online social groups could be classified as sharing views and experiences (55.5%), encouragement (28.7%), and knowledge and information (15.8%). About half of the participants' posts (52.9%) addressed the themes listed in the U.S. Clinical Practice Guideline for preventing smoking relapse. The participants perceived the posts as useful reminders for smoking cessation, but avoidance of reporting relapse, inactive discussions, and uninteresting content were barriers to the success of the intervention.

Discussion:

Online social groups provided a useful platform for the delivery of cessation support and encouragement of reporting abstinence, which support relapse prevention. The effectiveness of such intervention can be improved by encouraging more self-report of relapse, active discussions, sharing of interesting content, and using an appropriate discussion platform.

Conclusion:

Quitters who participate in the online social groups can benefit from peer support and information sharing, and hence prevent smoking relapse.

Introduction

Many smokers relapse after having recently quit due to withdrawal symptoms, smoking urges and triggers. In addition to traditional smoking cessation intervention, many online social groups have been developed on Facebook, Twitter, and Internet-based blogs to engage quitters to support their peers by sharing quitting information and experience.^1

–5 Content analyses of these groups have identified psychosocial and informational “cyber-support” for smoking cessation in an online network of smokers.^1
–3 Patterns of posting behavior,¹ classification of users,⁵ and the association between posting behavior and the effectiveness of smoking cessation³ have also been studied. However, no study has examined how discussion content reflects the intervention components listed in the smoking cessation intervention guidelines used by healthcare professionals. A systematic content analysis can be used to assess whether the discussion content corresponded to the behavioral support suggested by the clinical practice guidelines.⁶

Our pilot randomized controlled trial (RCT) was the first to show evidence of the effectiveness of relapse prevention for recent quitters through moderators' reminders and online discussion on social media.⁷ To understand the components of cyber-support within the online social groups that have contributed to the relapse prevention, a content analysis to classify the cyber-support and discussion topics is needed. Also, participants often disclosed specific needs and experiences in group discussion,^4,8 which may not be easily identified in surveys or experimental studies. Exploring these experiences from the discussion contents may help identify further improvement in the intervention.

We evaluated relapse prevention through WhatsApp and Facebook social groups by triangulating findings from the content analysis of online social groups with additional qualitative interviews. We had three specific research questions: (1) What cyber-support was delivered in the online social groups for recent quitters? (2) Did the discussion content include the components for helping recent quitters prevent relapse recommended in the U.S. Clinical Practice Guideline for Smoking Cessation? and (3) How did the recent quitters evaluate the effectiveness and feasibility of receiving cyber-support through online social groups?

Materials and Methods

Study Design

The original study was a pragmatic pilot RCT of smoking relapse prevention in a sample of recent quitters. This study included a qualitative analysis of discussion content from the online social groups of WhatsApp and Facebook, as well as qualitative interviews with the participants from that RCT. The study was approved by the Institutional Review Board of the University of Hong Kong/Hong Kong Authority Hong Kong West Cluster (IRB reference no. UW-13-528).

Participants and Data Collection

Details of the RCT and main results were given in our previously published report.⁷ Briefly, the participants were clients in the Tung Wah Group of Hospitals Integrated Centre of Smoking Cessation (ICSC) in Hong Kong, China, who had finished the free 8-week smoking cessation intervention, including counseling and medications. Eligible participants (n = 136) who reported no smoking in the past 7 days were invited to give a written consent to participate in the RCT and agree with their posts in the social groups being analyzed without disclosing personal information in the result dissemination. They were then randomly allocated to the three RCT groups (WhatsApp, Facebook, or Control group). Participants in the WhatsApp (n = 42) and Facebook (n = 40) groups were invited to join the online discussion through WhatsApp or Facebook, respectively.

As this was a pilot RCT with limited resources, the time period for the online group discussion was 8 weeks. They were encouraged to seek information on quitting and share their quitting experiences in Chinese. All participants in each recruitment week were allocated into one social group led by a moderator who was a trained smoking cessation counselor in ICSC. The moderator posted in the group thrice a week according to the moderator protocol. To facilitate peer discussion, the moderators participated in the discussion when no participants responded to others' posts for a day or when any participants needed advice from the moderator. All posts could be read only by the participants and the moderator in the social group. All discussion content was archived and converted to anonymous transcripts for content analysis.

Control group received only the same self-help booklet and were advised to contact ICSC's counselors when they experienced smoking desire or relapse due to any smoking cues (usual care). Tobacco abstinence and relapse were assessed by telephone interview, and validated by their exhaled carbon monoxide and salivary cotinine at 2- and 6-month follow-up.

After the intervention, we used purposive sampling to select 13 participants (five from the WhatsApp and eight from the Facebook social groups) of both sexes, all age groups (young adult, middle aged, and elderly), and all smoking status (quitter and smoker) in the 51 participants who eventually joined the social groups and were contacted at 6-month follow-up. This type of sampling helped us collect opinions from a small number of participants with various characteristics. The first author of the study (CYTD) conducted each 30- to 60-min interview with a trained research assistant. An interview guide with open-ended and iterative questions was used to probe for more experiences from the interviewees. The interview content was transcribed verbatim in Chinese for further analysis. Each participant was offered HK$300 cash (US$38.5) to compensate them for their travel expenses and time.

Coding of the Conversation Content

Thematic codes related to social support and smoking relapse were developed by looking for posts regarding sharing experiences, encouragement, perceptions, knowledge, and suggestions based on previous studies of online psychosocial support groups.⁹ Each participant's post was initially coded with one or several thematic codes. To maintain consistency within the coding process, all discussion content was coded again independently by two trained research assistants.

Next, two panel discussion sessions with all coauthors were held to resolve any inconsistencies in the interpretation of the discussion content and code assignment. The revised thematic codes were then entered into QSR Nvivo 10 as nodes in the program (Appendix A1). The first author and another trained research assistant used these nodes to independently code each participant's post with the Nvivo. Cohen's Kappa (κ) was calculated for each thematic code to assess the interrater agreement between the two coders. Most κ of the nodes achieved moderate (0.6–0.8) or strong (0.8 or above) agreement (Appendix A1). The definitions of nodes with κ below 0.6 were discussed and redefined in the third panel discussion. The first author either coded again with a revised definition or reallocated those posts to other codes.

The thematic codes were then further classified based on two dimensions: cyber-support and discussion topics pertaining to smoking relapse. Cyber-support included psychosocial and informational support with reference to previous studies on Twitter-based group discussions,^1,10 including three themes: (1) sharing views and experiences, (2) encouragement, and (3) knowledge and information. In addition, posts that were not coded to the above categories were classified as “other.” By Nvivo, the first author allocated all thematic codes into these four themes. Some thematic codes were allocated to multiple themes, such as medications and suggest methods of quitting (Appendix A1).

The discussion topics on smoking relapse were constructed based on the U.S. Clinical Practice Guideline, which suggested responses to five problems encountered by recent quitters⁶: (1) lack of support for cessation, (2) negative mood or depression, (3) strong prolonged withdrawal symptoms, (4) weight gain, and (5) smoking relapses. The second category was a subset of the third category,⁶ so they were combined into “craving and withdrawal symptoms.” The category “lack of support for cessation” in the guideline addressed only the support from existing healthcare and clinical services,⁶ excluding nonclinical sources of social support and other environmental factors. Hence, this category included thematic codes of self-encouragement, mutual encouragement, environmental smoking cues, ask questions about ICSC, and support from family or friends. Moreover, posts pertaining to reporting abstinence, medication, and alcohol were identified in the coding process and hence were included among the discussion topics.

Coding of the Qualitative Interview

We analyzed the qualitative interview transcripts using framework analysis¹¹ to construct a coherent and logical structure from the classification of many opinions and perceptions of social groups. The first author and two trained research assistants first used the open-coding method to identify relevant content in the transcripts; next, the first author classified all labeled content into several independent themes showing distinct features of the social groups. The analyzed results were then discussed and consolidated in the panel meetings with the coauthors.

Results

Participants and Postcharacteristics

The RCT started 7 WhatsApp and 6 Facebook social groups for a total of 42 and 40 participants, respectively (Table 1). The mean number of participants in each group was 6.3 (standard deviation 2.5; range 2–10). Seven participants in the WhatsApp groups and four participants in the Facebook groups were dropouts as they could not be added to the social groups or left the groups early.

Table 1.

Characteristics of the Social Group Participants

BASELINE CHARACTERISTICS	WHATSAPP, n = 42	FACEBOOK, n = 40	ALL N = 82
Gender, n (%)
Male	32 (76)	29 (73)	61 (74)
Female	10 (24)	11 (28)	21 (26)
Age (years), mean (SD)	40.4 (10.4)	37.6 (8.0)	40.5 (9.9)
Marital status, n (%)
Single	11 (26)	15 (38)	26 (32)
Married	24 (57)	23 (58)	47 (57)
Other	6 (14)	2 (5)	8 (10)
Missing	1 (2)	0 (0)	1 (1)
Monthly personal income (HK$), n (%)
<$10,000	4 (10)	8 (20)	12 (15)
$10,000–$19,999	17 (41)	16 (40)	33 (40)
$20,000–$29,999	12 (29)	6 (15)	18 (22)
≥$30,000	2 (5)	8 (20)	10 (12)
Missing	7 (17)	2 (5)	9 (11)
FTND,^a n (%)
Mild (Score 3 or below)	14 (33)	12 (30)	26 (32)
Moderate (Score 4–6)	14 (33)	18 (45)	32 (39)
Severe (Score 7 or above)	14 (33)	10 (25)	24 (29)
Any quit attempt before intake, n (%)	35 (83)	31 (78)	66 (81)
Daily cigarette consumption, mean (SD)	14.5 (6.3)	15.3 (6.5)	14.9 (6.4)

FTND = Fagerstrom Test score for Nicotine Dependence level.

SD, standard deviation.

During the 2-month intervention period, the mean (WhatsApp: 56.5; Facebook: 18.5) and median (WhatsApp: 45; Facebook: 6) number of posts per social group was greater in the WhatsApp groups than the Facebook groups (independent-samples t-test for mean: p-value = 0.16, median test: p = 0.10) (Table 2). Of the WhatsApp participants, only 7.1% (n = 3) did not post, more than half (54.8%, n = 23) posted 1–9 times, and 21.5% (n = 9) posted 10 times or more. The corresponding percentages for Facebook groups were 47.5% (n = 19), 35.5% (n = 14), and 7.5% (n = 3), which shows that they had apparently lower participation level than the WhatsApp counterparts. The number of moderators' posts on WhatsApp and Facebook was 465 and 255, respectively. Of all the moderators' posts, 43.3% (n = 312) were standardized posts listed in the moderator protocol and 56.7% (n = 408) were related to moderators' discussions with the participants.

Table 2.

Descriptive Statistics of the Postcharacteristics

POSTCHARACTERISTICS	WHATSAPP	FACEBOOK
No. of participants	42	40
No. of social groups	7	6
Participants per social group, range	2–9	4–10
Participants' posts
Total no.	356	111
Group median (range)	45 (5–133)	6 (0–84)
Group mean (SD)	56.5 (44.5)	18.5 (32.5)
No. of posts per participant, n (%)
Dropped out^a	7 (16.7)	4 (10.0)
Zero	3 (7.1)	19 (47.5)
1–9	23 (54.8)	14 (35.5)
10–19	7 (16.7)	2 (5.0)
≥20	2 (4.8)	1 (2.5)
Moderators' posts
Total no.	465	255
Median (range)	60 (34–85)	31.5 (20–109)
Mean (SD)	66.4 (18.3)	42.5 (37.5)

Drop-out participants included those who could not add themselves to the social groups due to technical problems or left the social groups early.

Cyber-Support

Of all the participants' posts (N = 467), 259 posts (55.5%) were coded as sharing views or experiences, 134 (28.7%) were encouragement, and 74 (15.8%) were knowledge and information. The other 88 posts (18.8%) were classified as others, as they were self-introductions, greetings, jokes, or expressions that were not directly relevant to cyber-support (Table 3). The distribution of cyber-support between the WhatsApp and Facebook was similar.

Table 3.

Quotes About Cyber-Support from the Social Groups

Sharing views or experiences (n = 259, 55.5%)

WhatsApp: 187 (52.5%); Facebook: 72 (64.9%)

• I felt better in my breathing and more energetic after quitting (Benefits of quitting)

• Cannot concentrate at work and irritated because of my craving to smoking. (Craving and withdrawal symptoms)

• I relapsed yesterday because my colleagues gave me cigarettes (Environmental smoking cues)

• I have changed from high doses to low doses of NRT, so I feel it tougher to remain abstinence. (NRT)

• My weight increased 4 kg last month. How about you guys? (Weight gain)

• I have quit for 3 months, and now have no desire to smoke (Reporting abstinence)

• I was determined to quit, but I relapsed as I had so much trouble in living … I have lost the hope to quit. (Reporting lapses)

Encouragement (n = 134, 28.7%)

WhatsApp: 107 (30.1%); Facebook: 27 (23.4%)

• Other can quit, why can't I?

• To those quitting and those who had quit: We must make all effort and quit smoking!

• I can quit even though I live alone, so you must be able to quit too!

Knowledge or information (n = 74, 15.8%)

WhatsApp: 41 (11.5%); Facebook: 33 (29.7%)

• I reduced smoking first and then used NRT to reduce craving.

• I exercised for 15–30 minutes a day, because a large tummy is problematic.

• I craved when I felt frustrated or busy. I would recall the benefits of quitting. The craving would then go away.

Others (n = 88, 18.8%)

WhatsApp: 72 (20.2%); Facebook: 16 (14.4%)

• Hi everyone! My name is Ah Man.

• Good morning!

• I am good. Haha.

• Thank you!

Remark: Each post could be coded with one or several thematic codes.

The category sharing views or experiences included diverse personal experiences of smoking cessation and participants' views on some controversial topics regarding smoking cessation. For example, they discussed whether smoking one cigarette would lead to a “full-blown” smoking relapse. Posts using cheering phrases, self-encouragement to increase confidence or commitment to quit, and showing determination and WhatsApp “Emoji” icons to respond to others' encouragement were classified as encouragement. The category knowledge and information included suggestions about quitting and weight control, medications (nicotine-replacement therapy [NRT]), and asking questions about ICSC.

Discussion of Smoking Relapse

About half of the posts (n = 247, 52.9%) were classified as one or several discussion topics about smoking relapse listed in the U.S. Clinical Practice Guideline, including lack of cessation support (26.3%, n = 123), craving and withdrawal symptoms (7.1%, n = 33), weight gain (5.1%, n = 24), (2.8%) smoking relapses (2.8%, n = 13), reporting abstinence (13.1%, n = 61), medications (6.6%, n = 31), and alcohol and smoking (2.8%, n = 13). The other 220 posts (47.1%) were not classified, as they were irrelevant to smoking relapse (Table 4). The distribution of the discussion topics between WhatsApp and Facebook is similar.

Table 4.

Discussion Topics on Smoking Relapse

DISCUSSION TOPIC^a	ALL N = 467, n (%)	WHATSAPP N = 356, n (%)	FACEBOOK N = 111, n (%)	QUOTES
1. Lack of cessation support^b	123 (26.3)	89 (25.0)	34 (30.6)	• My ICSC counselor was helpful. He listened to me and taught me a lot on quitting. (Existing services)
				• I live alone and I smoked a lot. I quit already, and I know you guys can do so! (Environmental support)
2. Craving and withdrawal symptoms	33 (7.1)	24 (6.7)	9 (8.1)	• I chewed gum whenever I had the craving.
				• I craved when I had a low mood.
				• I had serious coughing after I started to quit.
3. Weight gain^c	24 (5.1)	17 (4.8)	7 (2.0)	• After quitting for 3 months, my appetite increased. I always ate a lot at midnight.
				• I exercised for half an hour and practiced sit-up for at least 10 times. Also I ate as much fruit as I can.
4. Smoking relapses	13 (2.8)	13 (3.7)	0 (0.0)	• I secretly smoked half and then consumed the whole pack. Also, I smoked a cigarette when I drank alcohol.
				• I have forgotten to use the patch in the morning, so I relapsed in the afternoon… I admitted the failure, but I am still confident to quit again.
5. Reporting abstinence	61 (13.1)	40 (11.2)	21 (18.9)	• I have quit for 3 months already. Now I don't have any smoking desire.
				• I have quit for 2 months. I feel my well-being is getting better. Although I wanted smoking occasionally, I still can control it.
6. Medications	31 (6.6)	21 (5.9)	10 (2.8)	• I found I had forgotten to use the patch after having the craving for an hour.
				• I tried patch for 2 weeks but I felt very itchy and the skin got swollen. Then I stopped it and quit with my determination.
7. Alcohol and smoking	13 (2.8)	11 (3.1)	2 (1.8)	• I think smoking and alcohol was related, but my smoking was not triggered by alcohol.
				• I always smoked when I drank alcohol.
8. Irrelevant to smoking relapse^d	220 (47.1)	173 (48.6)	47 (42.3)	• I started to smoke because my girlfriend asked me to try it.
				• Good morning!

Each post could be coded with one or several topics.

Included self-encouragement, mutual encouragement, support from family or friends, ask questions about ICSC, and environmental smoking cues.

Included methods of weight control.

Included other irrelevant posts such as Emojis, introductions, greetings, jokes, and appreciations.

ICSC, Integrated Centre of Smoking Cessation.

Lack of cessation support included encouragement, support from family, friends, and ICSC counselors, and environmental smoking cues such as exposure to other smokers. Posts on craving and withdrawal symptoms included negative effect, lack of concentration, coughing, and restlessness. The most commonly reported solution was the use of NRT, in which the duration of NRT use, side effects, perceived effectiveness, attitude toward NRT, and using habits were shared. Those NRT-related posts were also coded as medication. Weight gain included experiences of weight gain due to smoking cessation. The participants did not regard weight gain as a serious health matter and a barrier for smoking cessation. Posts pertaining to smoking relapses (n = 13) primarily contained recent smoking relapses and the causes to which the posters attributed them. However, many more posts showed their abstinence, which were classified as reporting abstinence (n = 61), typically describing the length of abstinence so far, methods of quitting, encouragement, or how to combat environmental smoking cues.

Qualitative Interviews

The analysis of the 13 qualitative interview transcripts showed both positive and negative attitudes toward the online social groups. First, these groups provided useful reminders for maintaining abstinence (Appendix A2). Ten participants (WhatsApp: 5; Facebook: 5) perceived the moderator's posts, participants' sharing of methods, success stories, and encouragement as useful reminders for maintaining abstinence. Second, these groups encouraged the sharing of abstinence status and experience. Six participants (WhatsApp: 1; Facebook: 5) agreed that the online social group encouraged them to share their “successful” experiences with other quitters, which could demonstrate their abstinence to others and motivate others to quit. Only one participant refused to disclose his abstinence to other people because he thought that it was embarrassing if he relapsed again.

In contrast, five Facebook participants perceived smoking relapse as a failure. They felt uncomfortable posting their relapse experiences and assumed that the purpose of the social group was to share successful quitting experiences, not stories of smoking relapse. In addition, they were worried that their failure might trigger other people to smoke again or disappoint those people who had helped them.

Eleven participants (WhatsApp: 4; Facebook: 7) commented that the participation in the discussion was not active enough. They reported occasions of very few or even no responses from the participants or the moderator in the social groups. The participants suggested several reasons for the inactive participation, including uninteresting discussion content, too few participants in the group, feelings of embarrassment when sharing with strangers, and perceptions that the social group was intended for cessation reminders only. Some respondents reported that other Facebook newsfeeds were more interesting than the smoking cessation content in the online social groups and hence easily distracted their attention.

Discussion

Summary of Findings

We found that the participants in WhatsApp social groups posted more frequently than the Facebook counterparts, but the distributions in cyber-support and discussion topic were similar. Many participants' posts in the online social groups were relevant to cyber-support for smoking cessation by the sharing of views, experiences, and encouragement. Only 12% of the posts were about knowledge and information related to smoking cessation. About half of the participants' posts addressed the components for helping recent quitters prevent relapse recommended in the U.S. Clinical Practice Guideline for preventing smoking relapse,⁶ suggesting that the online social groups were useful platforms for relapse prevention, especially in the delivery of cessation support and encouragement of reporting abstinence. The participants perceived the posts as useful reminders for smoking cessation, but their avoidance of reporting relapses, inactive discussion, and perceptions of uninteresting content in the online social groups were barriers to successful intervention.

Interpretation

About half of the posts shared views or experiences and one fourth addressed the lack of cessation support through encouragement. These interactions may facilitate the development of trust and mutual support, which enabled the accumulation of emotional support.¹² Despite their superficial relationships, they nevertheless shared the common goal of smoking cessation. This connection was able to sustain mutual support, strengthening the view that computer-mediated communication with a common goal (e.g., relapse prevention) helps generate connections.^13,14 In addition, the private social groups used in our study may enhance online interaction without worrying on the leakage of posts. WhatsApp and Facebook were highly acceptable as they were the most commonly used online communication tools in Hong Kong, therefore might have increased accessibility and adherence. All these suggest that online social group discussion is suitable for the participants to establish emotional support on relapse prevention.

Social connection was also enhanced as the participants shared information and knowledge on craving and withdrawal management, weight control, and medications. Most importantly, their personal experiences might be more convincing and practical than expert advice on relapse prevention, supporting that online environment enhances the diffusion of ideas and information, particularly for networks of low emotional attachments and weak relationships.^12,15 However, we found more posts on emotional support than informational support. It was probably because the participants had received intensive smoking cessation intervention in the clinics and were abstinent when joining the social groups. They might need less informational assistance than unassisted quitters.⁷

Our qualitative interview found that the participants had different perceptions toward abstinence and relapse. Abstinence was perceived as success, achievement, happiness, and pride, whereas relapse was associated with failure, incapability, disgrace, embarrassment, and regret. Such a contrast in perceptions was consistent with our content analysis of the posts, which showed that abstinence was more frequently posted or discussed than relapse or withdrawal symptoms in our online social groups. This cyber behavior can be explained by the social identity theory that individuals strategize to achieve and maintain a positive identity in a social group.¹⁶ Abstinence as the positive identity had been reinforced through the group discussion, health education, and interventions. Conversely, the avoidance of sharing relapse experiences suggests a marginalization of this common postabstinence behavior. This cyber behavior may hinder early intervention and support, which are crucial to prevent a “full-blown” relapse and reengage relapsers in quitting.¹⁷ Hence, sharing of relapse experience and identification of smoking relapse through online social groups should be encouraged.

Despite strong evidence of the effectiveness of group-based behavioral interventions for smoking cessation,^18,19 engaging participants in online discussion is challenging. To increase engagement, as suggested by some participants, inclusion of interesting content and fun elements, and increasing the group size are needed. Also, we found that participants' advice and encouragement were convincing and appealing, because these experiences may be more accepted than expertise information. Our results showed that online social groups successfully crowdsourced the quitters' experience and information, echoing the findings of the previous study that peer messages were effective in increasing engagement in a web-assisted smoking cessation intervention.²⁰

Another factor to increase engagement of social groups is the discussion platform. We found more participants' posts and interactions in the WhatsApp than the Facebook social groups, which was in line with our previous findings on the effectiveness of WhatsApp.⁷ Therefore, participation level is important in affecting intervention effectiveness. Instant messaging tools like WhatsApp appeared to be more suitable for group-based intervention.

Limitations

Our study had a few limitations. First, all subjects were recent quitters who had received a smoking cessation intervention from community-based smoking cessation clinics. The extension of the intervention to quitters using unassisted smoking cessation warrants further investigation. Second, some posts appeared to be of ambiguous relevance to cyber-support. For instance, simple casual replies (e.g., “Thank you” or “Ok”) can be considered either causal relationship-building communication or polite responses intended to stop the conversation. This ancillary communication warrants further analysis. Third, the small number of qualitative interview (13 out of 51 participants) may affect the findings, and we could hardly have a balanced selection of participants in the WhatsApp and Facebook groups. Last, this study aimed to evaluate the cyber behavior of the recent quitters; hence, the moderators' posts were not coded. A content analysis of the moderators' posts would be useful to improve the practices of moderators.

Conclusions

This qualitative study showed that the emotional and informational support through online social groups addressed the themes of relapse prevention in the US Clinical Practice Guideline of smoking cessation. The participation level and effectiveness of the online group discussion can be further enhanced by encouraging more participants to report relapse and share their experience, and using an appropriate discussion platform.

Footnotes

Acknowledgments

We are grateful to the Tung Wah Group of Hospitals Integrated Centre on Smoking Cessation (ICSC) and Tobacco Control Office of Hong Kong Department of Health for funding this project and Ms Margaret Wong, Ms Jeanny Tam, and Dr Ho Kin Sun from ICSC for their suggestions in designing the intervention and evaluation. We thank Ms. Vivian Chan, Mr. Jonah Lai, Ms Rain Wan, Mr Alex Lau, and Ms Tina Fung for their contribution as social group moderators to deliver the interventions to the exsmokers. We thank Ms Zoe Wan and Ms Christina Lam for constructing the counseling content of relapse prevention. Finally, we thank the ICSC's counselors and HKU student helpers for their assistance in participant recruitment and data collection.

Disclosure Statement

The study was funded by Tung Wah Group of Hospitals Integrated Centre on Smoking Cessation, which was funded by Tobacco Control Office of Department of Health. Prof Tai-hing Lam is the principal investigator of the FAMILY project, which was funded by the Hong Kong Jockey Club Charities Trust. All other authors do not have connection with the tobacco, alcohol, pharmaceutical, or gaming industries, and nobody was substantially funded by one of these organizations.

Appendix A2.

Major Comments in the Qualitative Interviews

MAJOR COMMENTS	EXAMPLES OF QUOTES
Theme 1: Use of the online social groups
1. Reminders for maintaining abstinence	• I think the reminders and warnings were useful. It reminded me to stay abstinence.
	• I encouraged a man to quit in my group. He was a truck driver who traveled between Hong Kong and the China Mainland. He perceived quitting difficult. I wished he would succeed.
	• The good thing of using the social group was the posts could remind everyone to quit. Some people were inactive to respond, but everyone could at least read the posts.
2. Sharing of abstinence status and experience	• I have quit with the support from the ICSC, thus I want to help others too.
	• I used Facebook to share a video and received many “likes” in return.
	• As I have quit, I can tell the moderator that I succeed now … Personally, I like to “show off” in the WhatsApp group.
Theme 2: Barriers of participating in the social group discussion
3. Reluctance in reporting relapse	• Relapse is failure. It's meaningless to tell others I failed. I am afraid my post would affect other participants.
	• I felt embarrassed to tell others I relapsed again.
	• I didn't post frequently because I had not succeeded to quit. I can share my experience if I succeed. Other people may be more interested to know how I succeed. Otherwise, it would be better to keep silent.
4. Inactive discussion	• I only read the posts and had no response in the group. People can choose to respond or not in a social group, but they must do so in a face-to-face counseling.
	• My group had very few participants. They did not post at all. One participant only posted once.
	• The group was silent. I don't know why. The moderator kept encouraging the participants to respond. When they kept silent, the moderator could do nothing and had no idea what they were thinking.
5. Uninteresting content	• Some information was boring and inflexible because the quitting messages were not new to us. Quitting saves money, quitting is good to health etc. I think the posts' content need to cover more topics.
	• Firstly, the moderator was only concerned about the quitting status of the participants. The group was too much about smoking and quitting, so I just looked at the group posts for once or twice. I suggest organizing some group activities to appeal the participants.
	• The posts in the group could not draw my attention. Other Facebook newsfeed posts were more interesting than those in the social groups, and positioned at better positions in the newsfeed page.

ICSC, Integrated Centre of Smoking Cessation.

References

Rocheleau

, Sadasivam

, Baquis

, Stahl

, Kinney

, Houston

SLPTK

. An observational study of social and emotional support in smoking cessation Twitter accounts: Content analysis of tweets. J Med Internet Res, 2015; 17:e18.

Burri

, Baujard

, Etter

J-F

. A qualitative analysis of an internet discussion forum for recent ex-smokers. Nicotine Tob Res, 2006; 8(Suppl 1):S13–S19.

Myneni

, Fujimoto

, Cobb

, Cohen

. Content-driven analysis of an online community for smoking cessation: Integration of qualitative techniques, automated text analysis, and affiliation networks. Am J Public Health, 2015; 105:1206–1212.

Kurko

, Linden

, Kolstela

, Pietilä

, Airaksinen

. Is nicotine replacement therapy overvalued in smoking cessation? Analysis of smokers' and quitters' communication in social media. Health Expect, 2015; 18:2962–2977.

Healey

, Hoek

, Edwards

. Posting behaviour patterns in an online smoking cessation social network: Implications for intervention design and development. PLoS One, 2014; 9:e106603.

Fiore

, Jaén

, Baker

, et al. Treating tobacco use and dependence: 2008 update. Rockville, MD: Department of Health and Human Services, 2008.

Cheung

YTD

, Chan

CHH

, Lai

C-KJ

, et al. Using WhatsApp and Facebook online social groups for smoking relapse prevention for recent quitters: A pilot pragmatic cluster randomized controlled trial. J Med Internet Res, 2015; 17:e238.

Love

, Crook

, Thompson

, et al. Exploring psychosocial support online: A content analysis of messages in an adolescent and young adult cancer community. Cyberpsychol Behav Soc Netw, 2012; 15:555–559.

Willms

, Best

, Taylor

, et al. A systematic approach for using qualitative methods in primary prevention research. Med Anthropol Q, 1990; 4:391–409.

10.

Rui

, Chen

, Damiano

. Health organizations providing and seeking social support: A Twitter-based content analysis. Cyberpsychol Behav Soc Netw, 2013; 16:669–673.

11.

Ritchie

, Lewis

. Qualitative research practice: A guide for social science students and researchers. London, UK: SAGE Publication, 2003.

12.

Phua

. Participating in health issue-specific social networking sites to quit smoking: How does online social interconnectedness influence smoking cessation self-efficacy?. J Commun, 2013; 63:933–952.

13.

Walther

. Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Commun Res, 1996; 23:3–43.

14.

Tidwell

, Walther

. Computer-mediated communication effects on disclosure, impressions, and interpersonal evaluations: Getting to know one another a bit at a time. Hum Commun Res, 2002; 28:317–348.

15.

Ellison

, Steinfield

, Lampe

. The benefits of Facebook “Friends”: Social capital and college students' use of online social network sites. J Comput Mediat Comm, 2007; 12:1143–1168.

16.

Tajfel

, Turner

. The Social Identity Theory of Intergroup Behavior. Sidanius

JTJJ

(ed), Political psychology: Key readings, New York: Psychology Press. 2004.

17.

Marlatt

, George

. Relapse prevention: Introduction and overview of the model. Br J Addict, 1984; 79:261–273.

18.

Agboola

, McNeill

, Coleman

. A systematic review of the effectiveness of smoking relapse prevention interventions for abstinent smokers. Addiction, 2010; 105:1362–1380.

19.

Stead

, Lancaster

. Group behavior therapy programs for smoking cessation. Cochrane Database Syst Rev, 2005:CD001007.

20.

Coley

, Sadasivam

, Williams

, et al. Crowdsourced peer- versus expert-written smoking-cessation messages. Am J Prev Med, 45:543–550.