This study used a 2×2 factorial design (Table 1). Participants (N=34) were randomized to (1) varenicline + nicotine gum or lozenges (VAR+NRT; n=20) or varenicline alone (VAR; n=14) and (2) smartphone medication reminder messages (REM; n=11) or no reminder messages (NREM; n=23). Adaptive randomization [37,38] was used to assign participants to groups based on race, sex, cigarettes smoked per day, and education. Participants assigned to VAR+REM (n=2) received varenicline reminder prompts, and those assigned to VAR+NRT+REM (n=9) also received reminders to use gum or lozenges with decreasing frequency over time. Participants assigned to VAR+NREM (n=12) and VAR+NRT+NREM (n=11) did not receive medication reminders. Participants were followed for 27 weeks, beginning 1 week before their scheduled quit date through 26 weeks after their quit date, with key follow-up assessments of smoking status at 4, 8, 12, and 26 weeks post quit. Participants also completed daily smartphone assessments from 1 week before their quit date through 12 weeks after their quit date. All participants were enrolled between January 2020 and July 2021, and final 26-week follow-up visits were completed by February 2022. The study was discontinued early due to the Pfizer recall of varenicline (Chantix) following findings that nitrosamine levels were above the US Food and Drug Administration [39] acceptable intake limit. However, note that generic varenicline has become available since the conclusion of the study and can be accessed in future research.

Participants were asked to complete web-based assessments via the Research Electric Data Capture (REDCap; Vanderbilt University [40,41]) program at baseline (1 week before quitting) and then weekly through 4 weeks post quit with additional follow-up assessments at 8, 12, and 26 weeks post quit. Participants were compensated US $20 for each completed assessment (up to US $120).

Participants were prompted via the Insight smartphone app [42] to complete daily diary assessments 30 minutes after their self-reported usual wake time. Daily smoking, medication adherence, and other relevant variables were assessed from 1 week prior to the scheduled quit date through 12 weeks post quit. Participants also received medication reminder prompts (described in the Medication REM section). Those who responded that they did not take their medication as prescribed were asked about their reasons for nonadherence. Participants were compensated up to US $10 per week for the completion of daily diary assessments (up to US $130 total). Specifically, participants earned US $10 for completing 6-7 daily diary assessments or US $5 for completing 4-5 assessments each week. Participants who completed ≤3 daily diary assessments did not receive compensation for that week. The primary study outcomes were CO-confirmed, self-reported 7-day point prevalence abstinence at 4, 8, 12, and 26 weeks postquit follow-up.

Participants were initially asked to complete study follow-up assessments in person at the TTRP, but the onset of the COVID-19 pandemic led the university to limit in-person interactions on campus. As a result, between March and May 2020, all assessments were completed remotely via web-based assessments, daily smartphone assessments, and smartphone-based breath sample submissions. After the university reopened, participants were encouraged to attend study visits in person, but remote assessments were accommodated.

Approximately 1-week prior to the scheduled quit date, a Tobacco Treatment Specialist (TTS) provided an overview of the tobacco cessation program and assisted participants with developing a quit plan. Participants were instructed to quit smoking at bedtime or 10 PM (whichever occurred first) on the evening before their next weekly counseling session (≈1 week after enrollment). All participants were offered up to 5 additional weekly counseling sessions delivered in-person or by telephone with a TTS. Topics of discussion during the counseling sessions included (1) the impact of tobacco on the health benefits of quitting, (2) stress management strategies, (3) making positive lifestyle changes, (4) developing coping skills, and (5) relapse prevention. The TTS checked in with participants each week about the difficulties and successes they experienced and planned for anticipated future challenges.

Participants were evaluated by a study physician, and varenicline was prescribed. Doses were titrated during the first week (prequit) following enrollment (days 1-3: 0.5 mg once daily; days 4-7: 0.5 mg twice daily). Thereafter, participants were instructed to take 0.5 mg varenicline twice daily for 12 additional postquit weeks.

Participants randomly assigned to VAR + NRT were provided a supply of 2 or 4 mg nicotine lozenges or gum based on their prequit level of smoking (per package instructions) and their preference for gum or lozenges. Participants were instructed to start using gum or lozenges on their quit date and especially when they had the urge to smoke. Participants were initially instructed to use gum or lozenge every 1-2 hours during weeks 1-6 and to gradually reduce to every 2-4 hours during weeks 7-9 and 4-8 hours during weeks 10-12 after the scheduled quit date (per package instructions).

Participants randomly assigned to receive varenicline reminders received these reminders at times selected based on their preference (morning and evening after full titration), for example, “Don’t forget to take your Chantix pill this morning!” The reminders also inquired about whether the participant had taken their medication, for example, “You were scheduled to take your morning Chantix pill at 9 AM. Have you taken your Chantix pill this morning?” Participants were sent reminder messages twice before being asked about the main reason they did not take missed varenicline dose. Reminder prompts for varenicline began during the prequit period of the study.

Participants assigned to receive NRT reminders also received reminders to use NRT throughout the day, such as “Don't forget to use nicotine gum/lozenges regularly throughout the day! It is recommended that you use a piece of gum/lozenge every 1-2 hours.” They received up to 3 additional prompts to report whether or not they used a piece of gum or a lozenge. Participants received reminders every 3 hours from the scheduled quit day to 4 weeks post quit, every 4 hours during weeks 5-8, and every 6 hours during weeks 9-12 after the scheduled quit date. Note that reminder messages were delivered less frequently than the recommended dose of NRT (eg, every 1-2 hours during the first 6 weeks) due to concerns that the high frequency of reminder messages might be perceived as aversive or burdensome.

Participants self-reported their sex (male vs female), race (White vs racially minoritized [Black or African American, American Indian or Alaska Native, Asian or Pacific Islander, and multirace or other]), ethnicity (Hispanic vs non-Hispanic), years of age, education level (under high school vs under or above high school), and annual household income (<US $50,000 vs ≥US $50,000).

Participants reported the average number of cigarettes smoked per day prior to quitting and the number of years they had smoked at the time of enrollment.

The primary study outcomes were self-reported and biochemically verified 7-day point prevalence abstinence at 4, 8, 12, and 26 weeks after the scheduled quit date. CO was assessed using a portable Vitalograph ecolyzer (Vitalograph Inc; for in-clinic assessments) or via the Bedfont iCO Smokerlyzer (Bedfont Scientific Ltd; for remote smartphone-based assessments) with a CO level of ≤6 ppm indicating biochemical confirmation of self-reported abstinence (per current recommendations [35]).

Participants were considered abstinent each day if they did not report smoking during any smartphone-based reminder assessment on that day, and they reported that they did not smoke during the previous day on the morning daily diary. A variable reflecting the total number of days each participant was abstinent over the postquit period was created with a possible range of 0-84 days abstinent (ie, the first 12 weeks after the scheduled quit date).

Participants responded to the following smartphone daily diary question each morning and evening (depending on where they were in the titration schedule): “Have you taken your Chantix pill this [morning/evening]?” A dichotomous variable was created to indicate whether or not the participant had reported using varenicline as prescribed on a given day. For example, if a participant was supposed to take the medication once in the morning and once in the evening, but they reported only taking it once, they were considered nonadherent for that day. The total days of varenicline adherence were calculated with a possible range of 0-91 days (ie, 1 week before the quit date [titration period] through 12 weeks after the scheduled quit date).

During each morning daily diary, participants were asked via smartphone assessment, “How many pieces of nicotine gum or lozenges did you use yesterday?” The total pieces of NRT used were reflected as a continuous variable and were later dichotomized to <5 versus ≥5 pieces for each day.

The Medication Adherence Questionnaire (MAQ) [43] is a 4-item participant self-report of adherence to their smoking cessation medications during the previous week. Scores may range from 0 to 4 with higher scores indicating greater medication adherence. The MAQ was administered weekly from the scheduled quit date through 4 weeks after the scheduled quit date and again at the 8- and 12-week postquit follow-ups (7 assessments). Weekly scores were averaged across assessments to create an average adherence score. MAQ scores of 0 were assigned to those who informed study staff that they were dropping out of the study or discontinuing study medication during the weeks after they dropped out. Otherwise, a mean substitution for that week was used in place of missing MAQ scores. MAQ scores were dichotomized based on a median split to reflect higher versus lower adherence within the sample (<3.60 vs ≥3.60).

During smartphone assessments when participants reported that they had not taken their varenicline or NRT, they were prompted to select the main reason for missing their medication. Participants could choose from the following 7 responses: “I forgot to take it, I was away from home, I experienced side effects, I didn't feel like I needed it, I didn't think it is working, I have decided not to quit smoking,” or “other,” where they were able to write in their response. Similarly, participants who reported missing their NRT could choose from the following 9 options, “I didn't feel like I needed it, I forgot to take it, I didn’t have it with me, I don't think it is working, I ran out, I experienced side effects, I have decided not to quit smoking, I did not like the taste,” and “other,” which included a write-in option.

As shown in Table 3, at each of the study follow-ups, biochemically verified smoking abstinence rates for those assigned to VAR+NRT were higher than those assigned to VAR alone. Table 4 provides descriptive data on daily smartphone assessments and shows that participants assigned to VAR+NRT reported ≈3 more days of abstinence than those assigned to VAR over the first 12 weeks post quit.

At study follow-ups, biochemically verified smoking abstinence rates for those assigned to REM were higher across all follow-ups than those assigned to the NREM group (Table 3). Those assigned to REM reported ≈19 more days of abstinence than those assigned to NREM over the first 12 weeks post quit.

Participants assigned to VAR had higher average MAQ scores than those assigned to VAR+NRT (mean 3.52, SD 0.60 vs mean 2.97, SD 1.24). A higher proportion of those assigned to VAR reported high adherence on the MAQ compared with those assigned to VAR+NRT (Table 3). Smartphone-based assessments indicated the same pattern, with those assigned to VAR reporting ≈7 more days of varenicline adherence than those assigned to VAR+NRT (Table 4).

Participants assigned to REM had higher average MAQ scores than those assigned to NREM (mean 3.41, SD 0.83 vs mean 3.10, SD 1.14). A higher proportion of participants assigned to REM reported high adherence on the MAQ compared to those who did not receive reminders (Table 3). Smartphone-based daily diaries indicated that participants assigned to REM reported ≈18 more days of varenicline adherence than NREM (Table 4). In the subsample of participants assigned to VAR+NRT, those assigned to REM group reported using a slightly higher average number of pieces of NRT per day compared to those in the NREM group. Further, those assigned to REM reported more days where they used ≥5 pieces of NRT per day than those assigned to NREM (Table 4).

Of the specified reasons for missing varenicline doses (n=276 instances throughout the entire study period), the most frequently reported reasons were side effects (70/258, 27.1%) and forgetting (36/258, 13.9%). Of the participants who reported varenicline side effects, 75.2% (328/436) reported that side effects were mild or very mild, with participants most commonly reporting sleep problems (307/574, 53.5%) or nausea (172/574, 30%). Likewise, of the participants who indicated that they did not use any NRT on a given day (702 instances across all participants throughout the study period), the most common reasons were “I did not feel like I needed it” (206/668, 30.8%) and “I did not have any with me” (95/668, 14.2%).