Developing a Social Media Intervention to Connect Alaska Native People Who Smoke with Resources and Support to Quit Smoking: The Connecting Alaska Native Quit Study

Abstract

Background

Face-to-face tobacco cessation has had limited reach and efficacy in Alaska Native (AN) communities. We describe our two-phased approach to develop content for Connecting Alaska Native People to Quit Smoking, a Facebook group intervention to reduce barriers to evidence-based smoking cessation treatment for AN people in Alaska.

Methods

Phase 1 included semi-structured telephone interviews with 30 AN people who smoke and ten stakeholders. They provided feedback on existing content from the Centers for Disease Control and Prevention Tips campaign and AN digital stories. Phase 2 included an online survey with a new group of 40 AN smokers who provided feedback on existing content via a measure of perceived effectiveness and cultural relevance.

Results

Phase I results revealed participants evaluated content based upon story strength, relevance to AN culture, emotional appeal, relatability to AN people, and favorite video. No single posting was rated highly across all themes. All perceived effectiveness (PE) and cultural relevance median scores fell between 3.5 and 4.4 (range 1–5). PE scores varied across participant demographic groups.

Conclusions

Content embodying characteristics perceived to be most appealing, effective, and culturally relevant were selected for the private Facebook group content library with refinements made to incorporate images of AN people engaged in AN activities. PE scores indicate a need for a wide variety of content that moderators could pull from when conducting the intervention.

Implications

Social media content targeting specific population sectors, such as American Indian/AN people for tobacco cessation needs to be culturally tailored. Our approach provides a model others can follow to determine what is appealing, relevant, and effective messaging.

Clinical Trial Registration

NCT 03645941.

Introduction

Smoking is the largest preventable cause of morbidity, mortality, and excess health cost in the U.S.¹ In 2018, smoking prevalence was 14% among U.S. adults. At 23%, American Indian (AI) and Alaska Native (AN) people had the highest smoking prevalence of any U.S. racial/ethnic subgroup. Smoking prevalence among AN residents of Alaska was more than double that of Alaskan whites (42% vs. 17%).² However, tobacco cessation interventions specific to this population are limited.^3,4 Developing strategies to decrease AN tobacco use will contribute to the U.S. public health objective to reduce tobacco-caused health disparities,^5,6 and to the Healthy Alaskans 2020 target of decreasing smoking prevalence among AN adults.^2,7 Social media holds promise as a scalable, culturally tailored, intervention platform to promote engagement in cessation treatment for minority populations with shared cultural values, including AN people.

Our previous interventions utilizing intensive, face-to-face tobacco cessation counseling for AN pregnant women and youth were individually-based approaches with limited reach and efficacy.^8,9 Geographic remoteness, weather, and travel cost present barriers to providing to health care services for AN people in Alaska. Online social networks, such as Facebook (FB), the most utilized social media platform in the United States that is also widely used in Alaska, are potential tools for reaching, engaging, and connecting AN people in smoking cessation efforts¹⁰ with large reach at relatively low cost, representing a distinct advantage over face-to-face approaches. Research indicates internet-based interventions are effective for smoking cessation, but are associated with low utilization.^11–13 In contrast, highly interactive social media could increase the depth of engagement and connection with an extensive reach to underserved, diverse populations^14,15 using evidence-based content.

The Connecting Alaska Native People to Quit Smoking (CANQuit) study is a four-phase project designed to promote smoking treatment utilization and cessation, employing a culturally relevant, FB-delivered intervention. In the first two phases, we evaluate existing tobacco cessation content for perceived effectiveness and cultural relevance with AN smokers. In phase 3 we beta-test the Facebook intervention and our culturally tailored social media content with ten AN smokers. In phase 4 we will pilot test the Facebook intervention with 60 AN smokers. The beta and pilot test in Phases 3 and 4 will be moderated by AN tobacco treatment specialists. These moderators will post content for the Facebook intervention that is culturally relevant, adapting the storytelling approach used by the effective Centers for Disease Control and Prevention (CDC) Tips educational, national, general audience, mass media campaign.¹⁶ Based on factual health communication messaging, Tips features graphic, emotional, true stories told by former smokers to increase awareness of the smoking harms and encourage quitting. It has a specific call to action to use free, evidence-based state quitlines, and smoking cessation website resources (www.smokefree.gov). The campaign was effective for increasing quitline utilization and quit attempts on a population-level.^16,17 While the Tips campaign has not been adapted for nor evaluated among AI/AN people (only two of the 31 available stories feature an AI/AN former smoker), its personal story format promotes relatability and reduces the tendency for smokers to generate counterarguments or discount adverse health outcomes as uncommon because the stories feature real people – not actors.^16,17 Numerous studies, including prior work by our team,¹⁸ suggest that storytelling is culturally congruent, making the Tips format ideal for social media content development. Additionally, our Facebook intervention will promote current tobacco cessation resources that are available in Alaska. This includes the Alaska Quitline and Tribal health tobacco cessation programs.

Here, we report methods and results in the phases of our study dedicated to intervention development (phase 1) and refinement (phase 2).

Methods

CANQuit was approved by the Mayo Clinic and Alaska Area Institutional Review Boards and by the Alaska Native Tribal Health Consortium (ANTHC). Recruitment was conducted primarily via paid FB ads and in person at ANTHC clinics in outpatient quarters only. For phase 1, recruitment also specifically included the Yukon-Kuskokwim Delta Region, for which the Yukon-Kuskokwim Health Corporation Human Studies Committee granted approval. We registered the study with clinicaltrials.gov (NCT 03645941) and conducted participant enrollment and data collection for phases 1 and 2 between October 2018 and March 2019.

Health Communication Framework and Study Design

The overall four-phase CANQuit study design is described elsewhere.¹⁹ In the first two phases reported here, we used the cultural variance and surface/deep structure frameworks to tailor our health messages.^20–22 The cultural variance framework considers AN cultural influence on health behaviors including specific beliefs, norms (ie, communication styles and social acceptance of tobacco use), values, and AN knowledge systems. Surface structure matches materials and messages to observable social and behavioral characteristics (eg, AN people, clothing, and consumer preference). Deep structure incorporates cultural beliefs and values (eg, family, community and environment interdependence, and importance).

Phase 1: Qualitative Pre-Testing

Content

The goal of qualitative pre-testing was to evaluate the cultural fit, appeal, and efficacy of existing content from reputable sources and to seek input/advice on what AN people who smoke would prefer in a social media intervention. We evaluated 14 potential FB postings, consisting of 10 videos and four text/pictures. We acquired four videos from the CDC Tips education campaign,²³ each ranging from 0.5 to 1.33 min in length. One of the Tips videos was culturally tailored, featuring an AN person, while the other three were general audience. We acquired six videos from the ANTHC Marketing and Communication Department that were developed using a digital storytelling approach.^24–26 All six stories shared personal tobacco cessation stories, featured AN people, were culturally tailored, and averaged 2 min in duration.

Additionally, we obtained four general audience text/picture postings from the “30 Smoke-free Days” CDC FB event, held in 2017, one of which included an AN person. Postings included images combined with inspirational text celebrating quitting milestones. Similar to Tips videos, these postings included a call to action to contact the Quitline.

Participants

We recruited adult AN smokers statewide through targeted FB advertisements, word-of-mouth, and flyers posted in outpatient areas at the Alaska Native Medical Center in Anchorage and at Yukon-Kuskokwim Health Corporation community clinics. We used purposeful sampling²⁷ and stratified participants by demographic segment (sex, age group [19–29, 30–49, ≥50 years] and urban vs. rural residence). For the purposes of this study, rural residence included participants who live outside of the Anchorage area and participants residing in Anchorage were coded as “urban.” Following Krueger's²⁸ recommendation of 10–15 interviews per major subset we estimated a total of 40 interviews, equally split between men/women and urban/rural to achieve data saturation.

We screened candidates for eligibility via phone. Inclusion criteria included: self-reported AN race/ethnicity, current Alaska residence, and age ≥19 years. Participation required cigarettes to be the main tobacco product used and having smoked at least one cigarette per day during the past seven days. Because the planned intervention will promote cessation treatment and quitting, participants needed to consider or be willing to make a quit attempt. They also could not have been enrolled in a cessation program or have used cessation pharmacotherapy within the past 30 days. Finally, all participants had to have a FB account or be willing to create one.

Consistent with best practices for community engagement,^29,30 we chose to include a range of diverse stakeholders in the initial qualitative work, including potential end users and tobacco cessation professionals. The latter group was included with the goal of developing a sustainable intervention within the Alaska Tribal Health System. These individuals refer and provide tobacco cessation interventions for AN people and therefore have a stake in the health of AN people interested in stopping tobacco use. We interviewed 10 stakeholders who were Alaska residents comprising quitline coaches and tribal cessation program counselors, with some being AN people. These stakeholders, invited by ANTHC, provided their views on the proposed intervention content, with data saturation estimated at 10 interviews.

Procedures

Participants and stakeholders agreed to random assignment to one of two groups reviewing five out of 10 videos (to reduce respondent burden) and four text/picture postings, and to a follow-up interview. Participants could receive materials by email or mail. All interviews, lasting about 60 min, were conducted by one of two Alaska-based research staff. Both interviewers were trained on-site in Anchorage using didactic instruction and mock interview practices.

At the beginning of the interview we asked for demographic, health literacy,³¹ and subsistence lifestyle information. We assessed cigarettes smoked per day, time to first cigarette,³² readiness to quit (Contemplation Ladder score),³³ and use of other nicotine/tobacco products. Stakeholders were only asked demographic questions along with “How are you involved in helping AN people stop using tobacco?”

We developed a semi-structured interviewer guide that was pre-tested with research staff for duration, flow, and content. We provided participants and stakeholders an opportunity to explain whether video postings would be acceptable, effective, and resonate with AN culture and values by asking the following questions and probing where needed. We asked: What did you think about the story? What did you think of the person in the story? What parts of this video fit or don't fit with AN culture? How helpful do you think this video might be in getting an AN person to use a quit line or Tribal smoking cessation program? Which video was your favorite? What would you recommend to improve the videos?

For each text/picture posting, we asked: What did you think about the posting? What message did you take away from this posting? What did you think about the picture/text? How would you make this posting better? How might this post fit with AN culture? How do you think this posting might be effective in motivating an AN person to use a quit line or Tribal smoking cessation program? What makes you more likely to respond to a social media post?

Participants and stakeholders received a $25 Visa gift card for remuneration upon interview completion. All interviews were audio recorded, notes were taken by interviewers, and transcripts were generated. We reached data saturation for all interview topics, whereby no new information was being obtained.

Qualitative Analysis

QSR NVivo software, version 10 (Doncaster, Victoria, Australia) helped facilitate response theme generation, with codes and categories based on interviewer guide topics and themes emerging from the interviews. Two study team members coded responses together, discussing any coding discrepancies until reaching consensus,³⁴ consulting a third study member when necessary. In addition to open coding, we conducted planned comparisons within and across sex, age, and urban/rural strata, making connections between identified categories to achieve in-depth understanding of elements important to include as postings. Additionally, to aid in assessment of the “acceptability” of the video postings, we reviewed text related to each response by video and marked it as positive, negative, mixed (both positive and negative), or neutral in its content. We summarized our sample using descriptive statistics.

Phase 2: Quantitative Pre-Testing

For phase 2, we selected content (four videos and two text/image postings) to evaluate for perceived effectiveness based on Phase 1 input of their appeal and cultural saliency. As none of the content contained all the themes sought, we chose videos that might provide information on which themes to focus on for the intervention. For example, would a story that didn't feature an AN person, but did embody AN values, be perceived as effective? Or would a story that featured an AN person, embodied AN values, but didn’t include a smoker be perceived as effective?

Participants

We conducted a quantitative survey with a new sample of 40 AN adult smokers, using stratified purposeful sampling identical to Phase 1. We limited statewide recruitment to targeted FB ads and word-of-mouth, screening candidates by phone for eligibility criteria identical to Phase 1. Following verbal consent, we emailed each eligible candidate a link to the on-line survey. Participants received a $25 gift card upon survey completion.

Procedures

Respondents viewed each of the postings then completed a 20-min online Qualtrics (Provo, UT) survey assessing socio-demographic and tobacco use characteristics identical to Phase 1. We included an appraisal of perceived effectiveness (PE) for each of the postings, this 6-item validated measure has been used to evaluate the Tips stories and is similar to PE measures used in other smoking cessation research.^35–37 For PE, participants reported how much they agreed with the following six statements for each posting: “This was worth remembering,” “This grabbed my attention,” “This was powerful,” “This was informative,” “This was meaningful,” and “This was convincing” using a five point scale from 1 (strongly disagree) to 5 (strongly agree).

We produced a total PE score by summing scores for the six items and dividing by the number of items in the scale (ie, possible total score range 1 = strongly disagree to 5 = strongly agree); scores >3.0 indicated greater PE, consistent with the CDC Tips Evaluation.^35–37 For this study, we added “this fits with my culture” to assess cultural relevance for each posting, with the same response options and scoring.

Statistical Methods

We summarized sample characteristics, responses to the individual items, total PE score, and the cultural relevance item for each posting. The PE score was calculated as the mean of six items, producing a median calculated on non-whole numbers. We used the Kruskal Wallis test to analyze associations of the total PE score and cultural relevance item with sex, age group (<30, 30–49, and ≥50 years), and locality (urban, rural). P values ≤ .05 were considered statistically significant.

Results

We analyzed data February through April 2019.

Phase 1: Qualitative Pre-Testing

Participants

We interviewed 30 smoker participants and 10 stakeholder participants who were recruited between October 2018 and March 2019. Of 59 AN smokers screened, 58 met study eligibility criteria and gave verbal consent; of these, 30 completed the interview. Table 1 provides demographic and tobacco use characteristics of Phase 1 participants, who reported cigarettes smoked per day at 8.2 ± 4.6 (range 0–20) and their readiness to quit smoking was a 7.6 ± 2.7 (range 0–10) on a scale of 0 to 10 with 10 being most ready. Of 11 stakeholders agreeing to be contacted and interviewed, 10 completed the interview. Stakeholders consisted of Tobacco Treatment Specialists from regional Alaska Tribal health organizations and State of Alaska Department of Health and Social Services Tobacco Prevention Program employees.

Table 1.

Participant Socio-Demographic and Tobacco Use Characteristics: Phases 1 and 2^a

Characteristic	Phase 1	Phase 2
	N = 30	N = 40
	# (%)	# (%)
Gender
Men	8 (26.7)	15 (37.5)
Women	22 (73.3)	25 (62.5)
Locality
Rural	13 (43.3)	24 (60.0)
Urban	17 (56.7)	16 (40.0)
Age (years)
Mean ± SD	44.8 ± 11.9	41.1 ± 12.0
Range	26–70	21–65
19–49	20 (66.7)	29 (72.5)
≥50	10 (33.3)	11 (27.5)
Marital status
Single/divorced/widowed	19 (63.3)	35 (87.5)
Married/partner	11 (36.7)	5 (12.5)
Works for pay	12 (40.0)	14 (35.0)
Education
High school or less than high school	11 (36.7)	20 (50.0)
Some college/trade or technical/2-year or 4 year degree	28 (63.3)	20 (50.0)
Health literacy: “How often do you need or have someone else help you when you read instructions, pamphlets, or other written material from your doctor?” b
Never	26 (86.7)	32 (80.0)
Subsistence lifestyle: “What proportion of your food each year comes from hunting, fishing, and gathering (vs. store bought foods)?”
Less than half	24 (80.0)	20 (50.0)
About half or more than half	6 (20.0)	20 (50.0)
Cigarettes smoked per day
Mean ± SD	8.0 ± 4.6	7.1 ± 5.2
Range	0–20	1–24
Time to first cigarette ≤5 min	5 (16.7)	9 (26.5)
Contemplation Ladder scorec
Mean ± SD	7.6 ± 2.7	7.0 ± 2.5
Range	0–10	2–10
Use other tobacco/nicotine products	<5 (–)	8 (20.0)

^aSome response categories were collapsed (eg, for age group, health literacy, education) to protect confidentiality to suppress reporting numbers for cell sizes <5. This was done to protect participant confidentiality and minimize risk of exposing identities in this relatively small population.

^bOther response options were sometimes, often, and always.

^cPossible score range 0–10; higher scores indicate greater readiness to quit smoking.

Qualitative Interview Themes

Thematic categories for Phase 1 interview responses included: story and the storyteller, favorite video, recommendations for encouraging someone to respond to social media posts, and improvement suggestions. Within these categories, message relatability, family/community, AN culture/traditional lifestyle, effectiveness (ie, whether or not a post might be effective to prompt AN people to quit), emotional response, and positive or negative reactions emerged. The Supplementary Material provides illustrative quotes.

Story and the Storyteller.

When we asked how the content participants viewed fit or didn’t fit with AN culture, we found they felt more connected to the message when the storyteller was AN or stories featured an AN community with images and words alluding to AN culture, values, and experiences. Participants reported connecting emotionally and relating to images of family, children, and elders, as well as traditional activities, such as hunting, fishing, and snowmobiling. Alternatively, participants were more skeptical of stories that did not feature AN people, though they still related to the message if it fit with AN cultural norms. For example, when asked if they thought Bill's story fits with AN culture, one respondent said, “I thought he was going to be Native, but his story is real. A lot of our people are diabetic and do smoke. He was telling his story without sugar coating. I thought he was a good guy…” (f smoker 19–29, urban). Reactions to Terrie's story, containing dramatic images of a white woman dying from lung cancer, were mixed. One participant said, “She definitely was able to get the scare tactic message across and I heard those messages can be effective, but I'm not sure for Alaska Natives. She doesn’t look Alaska Native so I’m not sure it would be relatable. Her life seems very different than someone who grew up in the village smoking” (f smoker 19–29, rural). The issue of tobacco not being traditional to AN culture was raised by several stakeholder and smoker participants, specifically in reference to the Marie and Terrie videos. For example, a respondent said, “In my culture we didn’t have tobacco... It came from outside of our society. Native Americans down in the states grew tobacco. In Alaska no one grew tobacco until the white people brought it in” (f smoker ≥50, urban).

Favorite Video.

Participants selected Edith (n = 8, 28%, Grp. A), Marie (n = 8, 28%, Grp. A), Casandra (n = 10, 38%, Grp. B), and Tania (n = 8, 31%, Grp. B), all ANTHC digital stories told by an AN storyteller, as favorite videos. Similarly, stakeholders selected Edith (n = 2, 33%, Grp. A) and Casandra (n = 4, 67%, Grp. B) as their favorite videos. When looking at appeal across strata of age, sex, and urban/rural, Casandra and Edith had the most appeal from all groups. When asked why they selected these videos, participants most commonly reported the video contained AN people, AN scenes, and included experiences and reasons for quitting that resonated with them. One respondent to Casandra's story stated, “I don't want to be slowed down is what she said… the story was clear and concise, taking responsibility for your own actions is what I took away from that. That's what I remember is one of our family/traditional values, integrity” (m stakeholder 30–49, rural). Another participant said about Edith's story, “The way she told the story also really gets a person…My dad smoked three packs of cigarettes everyday…he died from a heart attack… It was very sincere” (f smoker ≥50, urban). Both stories included multiple images of the storyteller, enhancing relatability.

Improvement Recommendations.

For the videos and postings in which an AN person was not featured, including AN people was the most common recommendation for improving content. While participants liked many of the taglines of the posts, none of the postings featured AN persons. Participants recommended replacing pictures with images relating to AN culture/communities or featuring AN persons (surface structure). For videos, participants wanted to see the storyteller. Lastly, participants wanted to see a connection, such as a tagline, between the story and using a quitline or Tribal health system resource, as they noted was in the CDC content.

Recommendations to Prompt Responses to a Social Media Posts.

Participants indicated that posts that were emotional, inspiring, or funny would garner more responses. One participant commented, “When a person shares their story or images with engaging youth and activities to help people quit tobacco [to] draw me in or engage me.” Or when there is “Something I feel very strongly about or if I feel very strongly against. Something I find funny and relatable” (m stakeholder, 19–29, rural). A post including pictures and/or video would also increase responsiveness with participants more likely to click on pictures and respond to postings asking a direct question. Additional recommendations included using AN moderators for the Facebook beta and pilot test (phases 3 and 4) that participants can relate to (moderators of AN background who are able to understand and sympathize with participants because of their same cultural background), including quit coaches in the group, and providing incentives to participants for calling the quitline. Finally, several stakeholders commented that all the content needed to be altered to include a direct link to the Quitline, similar to the CDC content.

Moderator Postings Selected for Phase 2: Quantitative Pre-Testing

While we found all content to be both positively and negatively received, not one piece of content was considered “perfect” by any respondents when they evaluated it along the following themes: the strength of the story, its relevance to AN culture, its appeal to emotions, the relatability of the message, and what participants reported as their favorite video. While the AN digital stories were favorites, we were not sure how effective these would be if they included only some of the themes linked to being effective. For example, most participants said that Bill’s story was very relatable and would be effective. However, Bill is also white, which could lessen message effectiveness. In Phase 2, we needed to test whether a relatable story would trump the storyteller being non-Native.

Attempting to capture as many themes deemed important in Phase 1, we selected six postings for quantitative evaluation. We chose four videos featuring both men and women, from both ANTHC and Tips sources that were of variable length: Casandra and Tania (ANTHC), Michael and Bill (Tips) (Table 2). The two text/picture posts selected in Phase 1 represented an inspirational family theme and another an emotional/fear theme was “Always Remember Your Reasons for Quitting” and “Be Prepared. Your Lung Cancer Can Spread to Your Brain.” These postings also contained some themes not associated with effectiveness (ie, non-AN person and image depicting a non-AN form of fishing). The goal of testing these concepts in Phase 2 was to gather additional information on how to select postings that would provide maximum effectiveness for the intervention given that none of the postings were considered “perfect.”

Table 2.

Perceived Effectiveness Items and Total Score: Quantitative Pre-Testing, Phase 2 (N = 40)^a

	Moderator postings
	Videos				Text/picture
Perceived effectiveness	Casandra	Bill	Tania	Michael	Always remember your reasons for quitting	Be Prepared
	(cancer, dad quit, quit for husband/son)	(diabetes, lost leg)	(community health aide, role model)	(Alaska Native, has COPD)
Item	Mean ± SD	Mean ± SD	Mean ± SD	Mean ± SD	Mean ± SD	Mean ± SD
	Median	Median	Median	Median	Median	Median
1. This was worth remembering	3.82 ± 1.41	4.32 ± 0.89	4.07 ± 0.97	4.35 ± 0.86	4.08 ± 1.13	4.17 ± 0.75
	4.0	4.0	4.0	4.5	4.0	4.0
2. This grabbed my attention	3.95 ± 1.12	4.18 ± 1.01	3.98 ± 0.97	4.25 ± 0.84	3.37 ± 1.27	4.20 ± 0.82
	4.0	4.0	4.0	4.0	4.0	4.0
3. This was powerful	3.98 ± 1.10	4.27 ± 0.90	3.98 ± 0.92	4.27 ± 0.85	3.82 ± 1.21	4.05 ± 0.90
	4.0	4.0	4.0	4.0	4.0	4.0
4. This was informative	4.08 ± 0.98	4.13 ± 0.91	3.98 ± 0.83	4.27 ± 0.78	3.92 ± 1.01	4.23 ± 0.80
	4.0	4.0	4.0	4.0	4.0	4.0
5. This was meaningful	4.00 ± 1.03	4.27 ± 0.82	4.08 ± 0.83	4.27 ± 0.82	4.10 ± 1.07	4.20 ± 0.82
	4.0	4.0	4.0	4.0	4.0	4.0
6. This was convincing	3.85 ± 1.18	4.13 ± 1.07	4.10 ± 0.85	4.30 ± 0.85	3.92 ± 1.07	4.13 ± 0.83
	4.0	4.0	4.0	4.0	4.0	4.0
Total scorea	4.0 ± 1.0	4.2 ± 0.9	4.0 ± 0.8	4.3 ± 0.8	3.9 ± 1.0	4.2 ± 0.7
	4.2	4.3	4.0	4.3	4.0	4.1
7. This fits with my culture	3.9 ± 1.1	3.5 ± 1.2	4.0 ± 0.8	4.4 ± 0.6	3.7 ± 1.2	3.9 ± 1.0
	4.0	4.0	4.0	4.0	4.0	4.0

Item scores range from 1.0 (strongly disagree) to 5.0 (strongly agree). The observed item score range for all postings was 1.0–5.0.

COPD = chronic obstructive pulmonary disease.

^aBased on six items (#1–6), used to evaluate the Centers for Disease Control and Prevention (CDC) Tips videos and advertisements. Davis et al.³⁴ the total score median is the median of means, frequently resulting in a non-whole number.

Phase 2: Quantitative Pre-Testing

Participants

We recruited 44 AN smokers during March through April 2019. Of the 44 screened, 43 met eligibility criteria; of these 43, 40 enrolled and completed the survey. Table 1 provides demographic and tobacco use characteristics of Phase 2 participants, who reported cigarettes smoked per day at 7.1 ± 5.2 (range 0–20) and their readiness to quit smoking was a 7.0 ± 2.5 (range 0–10) on a scale of 0 to 10 with 10 being most ready.

Perceived Effectiveness Ratings and Their Correlates

All postings scored relatively high on ratings of PE and cultural relevance, with mean scores ranging from 3.5 to 4.4 out of a maximum score of 5.0 (see Table 2). Analyzing possible associations between PE and participant age group, sex, and residence, we identified the median score of all participant means for each posting. The median PE score of Casandra’s video was significantly (p = .027) higher for participants in rural (median 4.3) versus urban (median 3.9) locations. We also found PE differences based on locality for Michael’s video (median 4.6 rural vs. 4.0 urban, p = .029) and for the text/picture posting “Always Remember” (median 4.5 rural vs. 3.6 urban, p = .016). When comparing differences between men and women, the “Always Remember” posting had a PE score was significantly higher for men (median 4.8) compared to women (median 3.8; p = 0.005). For Tania’s video, PE scores were significantly higher for the older two age groups (30–49 years median 4.1 vs. ≥50 years 4.0) than in the youngest age group (<30 years median 3.7; p = .005).

In terms of cultural relevance, the median score was significantly higher for men than for woman for Bill's video (4.0 vs. 3.0, p = .002) and for the “Always Remember” posting (4.5 vs.3.0, p = .019). We detected no other statistically significant associations between PE or cultural relevance and participant characteristics.

Discussion

Our study provides novel information on developing a social media intervention for AN smokers statewide and provided us needed information on which content to include and adapt for the CAN Quit content library which will then be beta-tested (phase 3) and utilized by moderators for the pilot randomized controlled trial (phase 4). We solicited input from AN smokers and stakeholders on FB content to be tested in the ensuing intervention phases of this study. In Phase 1 we tested the content qualitatively through structured interviews and in Phase 2 we quantitatively tested six pieces of content selected from Phase 1. Participants preferred content featuring AN stories, people, families, and values. Evaluating the content for PE and cultural relevance, we observed differences in PE ratings between participants who lived in rural and urban communities and between sexes. PE was rated higher by participants living in rural compared to urban communities for Casandra's video, Michael's video, and the “Always Remember” posting. Casandra's video and the “Always Remember” posting featured individuals from rural Alaska and photos of rural Alaska, suggesting that people from rural communities preferred content that featured people and places they can relate to. Although Michael's story didn't state what type of community he comes from, he is of AN heritage, which may have made him more relatable to participants from rural communities. Compared to women, men rated Bill's video and the “Always Remember” posting higher in cultural relevance, both content that also featured men, suggesting that men prefer to see content they can relate to. The PE of the four videos (two CDC Tips Campaign videos and two ANTHC digital stories) we tested with AN smokers were rated between 4.0 and 4.3 (out of a possible 5.0). This compares closely to the CDC Tips Campaign videos with U.S. smokers with an eigenvalue of 4.4, with a positive reaction to videos across ethnicity irrespective of the race/ethnicity of the ad participant. Study limitations include small numbers of participants from whom we gathered data for this preliminary mixed methods pilot study. Thus, survey measures of PE and cultural relevance by participants in Phases 1 and 2 may not apply to all AN smokers interested in quitting tobacco use or willing to set a quit date. Additionally, there are more than 225 federally recognized AN tribes in Alaska who share common values of storytelling, family, community, and leisure activities. Thus, it would not be feasible to stratify by specific tribe. To ameliorate major potential differences, we collected data from a purposeful sample stratified by age, sex, and rural/urban status. Another limitation concerns Tribal health organizations that differ in the types of cessation resources and treatments they provide. We sought input from quitline coaches and cessation program counselors statewide so that referrals to the quitline and information specific to tribal health organizations resources were accurate and complete in all Alaska regions.

This study capitalizes on the popularity of social media platforms, such as FB, among AN people, and addresses three of seven gaps and priorities in tobacco research identified by the Tobacco Control Research Priorities Working Group³⁸ by working with an underserved population with tobacco use disparities, developing a novel population-specific behavioral intervention, and promoting uptake of evidence-based tobacco treatment and cessation. This pilot study is the first step to an intervention that could ultimately be widely disseminated through social media, internet, and mass media channels. If effective in advancing accessibility to tobacco cessation treatment for AN people in Alaska, where smoking prevalence remains high, the intervention could serve as a model for interventions in other populations such as AI smokers, enhancing overall reach and impact. The process of adapting the effective CDC TIPs content so it fits with AN cultural values is novel as is our adaptation of ANTHC digital stories to include the CDC format of a call to action and a Quitline link. Our results contribute relevant information to tobacco control research efforts to understand and apply effective risk communication strategies via social media, an important tool to maximize that could be replicated in future dissemination efforts.

Supplementary Material

A Contributorship Form detailing each author’s specific involvement with this content, as well as any supplementary data, are available online at https://academic.oup.com/ntr.

Funding

The research reported in this publication was supported by the National Institute on Drug Abuse (NIDA) of the National Institutes of Health under Award Number R34DA046008 (Patten). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Disclosures

Unrelated to this project, Dr. Prochaska has provided consultation to pharmaceutical and technology companies that make medications and other treatments for quitting smoking and has received funding from Facebook for planning evaluation of a mobile health intervention. Dr. Prochaska also has served as an expert witness in lawsuits against tobacco companies.

Declaration of Interests

All other authors have no conflicts of interest to disclose.