Article Text
PDF
Abstract
Introduction Widespread misperceptions of the health risks of nicotine could undermine the public health benefits of the Food and Drug Administration’s actions, including modified risk tobacco product authorisations and a reduced nicotine product standard for cigarettes.
Methods 794 US adults (aged 18+) in NORC’s AmeriSpeak panel participated in a randomised controlled trial in Spring 2021 to test the effect of three exposures to eight nicotine corrective messages (NCM) on beliefs about nicotine, nicotine replacement therapy (NRT), e-cigarettes and reduced nicotine content (RNC) cigarettes at 3-month follow-up. Analyses conducted in 2022 examined the effect of study condition (NCM (n=393) vs no message control (n=401)) on nicotine beliefs, use intentions and use of nicotine and tobacco products.
Results Exposure to three NCM doses reduced nicotine (b=−0.33; 95% CI −0.60, –0.07), NRT (b=−0.49; 95% CI −0.85, –0.14), e-cigarette (b=−0.32; 95% CI −0.59, –0.05) and RNC cigarette false beliefs (b=−0.64; 95% CI −1.26, –0.02) compared with the control, controlling for baseline beliefs. Baseline tobacco use and concern about nicotine addiction attenuated intervention effects on false beliefs about RNC cigarettes. There were few intervention effects on intention or use of nicotine and tobacco products.
Conclusions Repeated exposure to NCM was necessary to reduce false beliefs about nicotine and tobacco products. Future studies will improve understanding of the dose and duration of nicotine education needed to shift intentions and behaviour, as well as tailored content for tobacco product users to achieve similar reductions in false beliefs as non-users.
Trial registration number NCT04805515.
- Nicotine
- Public policy
- Public opinion
Data availability statement
Data are available upon reasonable request. Data are available upon reasonable request. The corresponding author will make deidentified participant data and the data dictionary available following publication. Institutions and individuals wishing to access any resources or data must contact the corresponding author (andrea.villanti@rutgers.edu). Data will only be made available to those whose proposed use of the data has been approved by the corresponding author. The recipient must agree to not transfer the data to other users and that the data are only to be used for research purposes. The investigators will require requestors of data to sign a data sharing agreement that will ensure (1) use of the data is only for research purposes, (2) data security using appropriate technology/firewalls, (3) destruction of data after data analysis and (4) proper citation in publications or other written materials. A record of transfer of data and a copy of the dataset that was distributed will be kept by Rutgers University.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Exposure to brief nicotine education may reduce false beliefs about nicotine. However, existing studies on nicotine education messaging have been limited to single-exposure studies and only two have experimentally tested a series of messages that might be delivered in a mass media health communication campaign.
WHAT THIS STUDY ADDS
Findings from this randomised controlled trial in US adults show that three exposures to the same series of nicotine corrective messages over 3 months reduced adults’ false beliefs about nicotine, nicotine replacement therapy, e-cigarettes and reduced nicotine content (RNC) cigarettes at follow-up. Past 30-day tobacco use and concern about addiction to nicotine at baseline attenuated intervention effects related to RNC false cigarette beliefs at 3-month follow-up.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Results suggest that comprehensive public education on nicotine that follows the Centers for Disease Control and Prevention’s recommended duration for a widespread public education effort can reduce false beliefs about nicotine, nicotine replacement therapy, e-cigarettes and RNC cigarettes in a general population sample of adults. These findings are relevant and timely given the Food and Drug Administration’s (FDA) authorisation of modified risk tobacco products and impending FDA rulemaking on nicotine content in cigarettes and other combusted products.
Introduction
Tobacco-related harms ultimately result from addiction to the nicotine in tobacco products1 and in June 2022, the US Food and Drug Administration (FDA) announced its intention to propose a rule restricting the maximum nicotine content in cigarettes in 2023.2 However, widespread misperceptions about the harms of nicotine3–6 have only increased in the USA over time.7 As of 2019, 86% of US adults understand that nicotine increases the appeal of tobacco products, but only 22% understand that nicotine does not ‘cause most of the cancer caused by smoking’.7 Communicating about the risks of smoking separately from the risks of nicotine will be essential to maximising the public health benefit of products authorised as modified risk tobacco products (MRTPs) and planned regulations to reduce the nicotine content in cigarettes.
Previous qualitative work has identified that people who smoke lack knowledge on nicotine separate from smoking,8 and even when they know the risks of nicotine they are inconsistent in their assessment of the relative harmfulness of tobacco products.9 Findings from our prior work suggest that exposure to a brief nicotine corrective messaging (NCM) intervention reduced nicotine misperceptions in convenience samples of adult tobacco users and non-users10 and adults who use opioids,11 a group with a high prevalence of cigarette use. The few studies that have tested candidate educational messages about reduced nicotine content (RNC) cigarettes have had mixed effects on harm perceptions.12–16 Byron and colleagues found that a message that conveyed the percentage reduction in nicotine content—similar to VLN’s ‘95% less nicotine’ claim—increased accuracy of beliefs about nicotine content and addictiveness of RNC cigarettes, but decreased accuracy of beliefs about cancer risk compared with a control message.15 Differding and colleagues found that a message about mortality (eg, ‘Cigarettes with 95% less nicotine are AS DEADLY as current cigarettes’) increased accuracy of beliefs about RNC cancer harms in people who currently smoke cigarettes compared with a control message, but not among people who do not smoke.16 These experimental studies, as well as a focus group study examining candidate messages about RNC cigarettes, underscore the difficulty in achieving accurate nicotine and RNC beliefs among people who do and do not smoke with the same messages.12 To date, existing studies on nicotine education messaging have been limited to single-exposure studies10 11 15 16 and only two have experimentally tested a series of messages that might be delivered in a mass media health communication campaign.10 11
Guided by the Theory of Planned Behavior,17 18 this study examined the effect of multiple exposures to a series of evidence-based nicotine corrective messages (NCM) on nicotine beliefs in a general sample of adults over a 12-week period. We hypothesised that adults in the NCM condition would report fewer false beliefs about nicotine, nicotine replacement therapy (NRT), e-cigarettes and RNC cigarettes at follow-up compared with those in the control condition. We also hypothesised that current smoking would moderate the effect of NCM on false beliefs of nicotine, NRT, e-cigarettes and RNC cigarettes, with the effect of NCM on these outcomes attenuated in adult smokers compared with non-smokers. Our secondary outcomes were intention and use of nicotine and tobacco products at follow-up.
Methods
Study overview
This study examines the effect of multiple exposures to a brief NCM intervention on nicotine, NRT, e-cigarette and RNC cigarette-related beliefs, and likelihood of subsequent tobacco and nicotine use in a population sample of adults. Sample size calculations were based on the effects observed for nicotine and NRT false belief outcomes in the pilot study.10 With 715 participants (~357 per group) and assuming 70% retention at 3-month follow-up, the study was powered to detect differences of <1 unit between the intervention and control group means for the nicotine, NRT and e-cigarette false belief scales and a 1.64 difference in the group means for the RNC cigarette false beliefs scale with 80% power and two-tailed alpha=0.05. This trial was registered on ClinicalTrials.gov (NCT04805515).
Participants
Participants were 794 US English-speaking adults aged 18 and older recruited in Spring 2021 from the NORC’s AmeriSpeak national consumer market research panel. The panel provides sample coverage of approximately 97% of the US household population.
Recruitment and enrolment
AmeriSpeak panel members eligible for the current study were English-speaking adults aged 18+ who were sent a standard email invitation describing a study about ‘Feelings on nicotine’. The email invitation introduced study procedures and directed interested panel members to the eligibility screener. Potential participants were identified by randomly selecting households within the panel with an adult within the study age range, and targeting study invitations based on recent study participation, and other demographic characteristics as needed to ensure a representative sample. If a given panel household had more than one active adult panel member, one adult was randomly selected for eligibility. Interested panel members who met eligibility criteria were directed to the online experiment.
Intervention and retention
Intervention
The nicotine corrective messaging (NCM) intervention condition was based on messages tested in our team’s pilot study.10 It included our six original messages and two new messages addressing nicotine in tobacco and e-cigarette products that were adapted from several evidence-based sources to be more accessible to a lay audience (online supplemental figure S1). The sources consisted of FDA’s 2017 comprehensive plan for tobacco and nicotine regulation,1 FDA’s 2013 modifications to labelling of NRT products for over-the-counter human use,19 the 2014 US Surgeon General’s Report on the Health Consequences of Smoking,20 reports on carcinogens from the International Agency for Research on Cancer21–23 and the National Academies of Science, Engineering, and Medicine report on the ‘Public Health Consequences of E-cigarettes’.24 Participants in the NCM condition were exposed to all eight messages in the same order at each exposure: (1) ‘Nicotine is the addictive substance in tobacco products’; (2) ‘Nicotine makes it easier for people to start smoking regularly’; (3) ‘Nicotine makes it harder for people to quit smoking’; (4) ‘Nicotine does not cause cancer’; (5) ‘Chemicals in cigarette smoke, not nicotine, largely cause cancer, heart disease, and other health problems in people who smoke’; (6) ‘Nicotine can be used safely in quit smoking products like nicotine patches, gum, or lozenges, even long-term’; (7) ‘E-cigarettes may expose users to significantly lower amounts of toxic substances than regular cigarettes. But they contain as much or more nicotine’; and (8) ‘Low nicotine cigarettes are as harmful as regular cigarettes. But they may help people quit and prevent new users from becoming addicted to cigarettes’. The Flesch-Kincaid Reading Grade level across all messages was 9.5.
Supplemental material
Exposure to the NCM intervention was assessed in two ways: (1) assignment to the NCM study condition or control and (2) number of times exposed to study messages, with control participants coded as 0 and NCM participants coded as 1, 2 or 3, depending on survey waves completed.
Study design
This study consisted of four survey waves that occurred over 3 months between 12 February 2021 and 14 May 2021. We allowed responses over 3 weeks to maximise the collection of wave 1 (weeks 1–3) and wave 4 (weeks 11–13) data and 2 weeks for wave 2 (weeks 5–6) and wave 3 (weeks 8–9).
In wave 1 of the main study, panellists completed baseline survey measures of nicotine beliefs and intentions/use of nicotine and tobacco products. They were then randomised in a 1:1 ratio within the survey system to the NCM intervention condition (n=393) or the delayed message control condition (n=401), with those in the NCM condition then receiving their first exposure to the study messages. NCM participants viewed each of the eight NCM messages for at least 5 s and were assessed for visual attention using a heatmapping task25 26 during which they were instructed to point and click at up to three areas of the message that attracted their attention. Panel members who completed the wave 1 survey were eligible to complete the remaining two surveys in waves 2 and 4, with panel members assigned to the NCM condition also invited to the wave 3 message exposure.
In wave 2, all participants completed a survey of nicotine beliefs and intentions/use of nicotine and tobacco products, which provided the first postexposure measures of the key outcomes. Participants in the NCM condition received their second exposure to the study messages and completed a heatmapping task.25 Only participants in the NCM condition received the wave 3 survey, which comprised the third exposure to study messages and the heatmapping task. At waves 1–3, control participants did not see any study messages. The wave 4 survey included the final assessment of nicotine beliefs and intentions/use of nicotine and tobacco products for the full sample; after collection of wave 4 outcome measures, all participants were exposed to the study messages and heatmapping task which served as delayed exposure for the control group. At the end of this survey, both conditions received a thank you message with a list of tobacco cessation resources. All surveys and research procedures were administered by fully trained and experienced NORC staff under the direction of ACV.
Retention
To encourage participation, NORC sent up to four email reminders per survey wave to panellists who had not yet completed the survey. This study also incorporated an incentive structure where participants received a $3 cash equivalent for completing the wave 1 and 2 surveys, a $2 cash equivalent for completing the wave 3 exposure and a $4 cash equivalent for completing the wave 4 survey. These incentives were administered by NORC through their standard procedures for panel members’ survey participation and are consistent with our team’s work with similar consumer research panels.
Measures
Outcomes
Our primary outcome variable was nicotine beliefs, using items from the National Cancer Institute’s 2017 Health Information National Trend Survey (HINTS) and earlier studies.3 27 28 These items were adapted and validated for use in this study5 10 29–34 and related specifically to nicotine’s addictive nature, nicotine as a cause of cancer and the role of nicotine in the health harms and cancer caused by smoking. In addition to assessing beliefs about nicotine, we also assessed NRT beliefs (6 items), e-cigarette beliefs (4 items) and RNC cigarette beliefs (11 items). These product-specific constructs (ie, NRT, e-cigarette, RNC cigarette beliefs) included items related to the addictiveness of the product and the perceived health risks of the product compared with regular cigarettes.
Secondary outcomes included intention and use of nicotine and tobacco products. Four items assessed intention to use cigarettes, e-cigarettes, nicotine replacement products or low nicotine content cigarettes in the next 12 months using items adapted from the PhenX Toolkit (710302). Ever use and number of days used in the past 30 days was asked for cigarettes, cigars, little cigars/cigarillos, e-cigarettes/vapes, smokeless tobacco, hookah/shisha/waterpipe, nicotine pouches and nicotine gum, patches or lozenges.
Baseline measures
Sociodemographics (eg, age, gender, race/ethnicity, education, employment, household income, region, metropolitan residence and internet access) were collected as profile variables by NORC. We assessed literacy using a validated, single-item measure,35 as well as concern about addiction to nicotine, smoking-related beliefs and cancer risk beliefs using measures from the 2017 HINTS.
Analyses
At the conclusion of data collection, NORC provided a deidentified data file of survey results and demographic data as well as full documentation of study procedures, including study invitation processes, recruitment and enrolment, and response rates across the four waves. First, we explored potential differences in effects of the NCM condition by baseline variables using bivariate analyses, then estimated differences between groups on wave 4 outcomes. Using multiple linear regression, we modelled the effects of study condition on change in the four false belief scales (ie, nicotine, NRT, e-cigarette and RNC cigarette), intention to use nicotine or tobacco products in the next 12 months and total days of use of nicotine and tobacco products in the past 30 days at waves 2 and 4. These analyses compared intent-to-treat analysis to per protocol analysis and complete case analysis. Prespecified moderators included past 30-day tobacco use at baseline, age group, sex and literacy; concern about addiction to nicotine was exploratory. We examined potential moderators of the relationship between study condition and wave 4 false belief scales using per protocol analysis and controlling for baseline false beliefs. Finally, we explored the role of exposure dose (number of times exposed to NCM) on wave 4 false belief scales using per protocol analysis.
Results
Of 2384 invited panellists, 794 or 31.6% of sampled members were eligible and completed the wave 1 survey and were randomised to the NCM intervention (n=393) or control group (n=401; figure 1). At wave 4, six hundred and nine of the 792 invited panellists (76.9%; n=290 NCM and n=319 control) completed the survey, with 551 completing all waves (complete cases). Half of participants were female and 62% identified as white, non-Hispanic; most (78%) reported having completed some college or more education, 82% lived in a metropolitan area and 90% had internet access in their household (table 1 and online supplemental table S1). Approximately 15% of participants reported past 30-day cigarette use. There were no baseline differences in participant demographics, tobacco use characteristics, literacy, nicotine/smoking beliefs or cancer beliefs by study condition; similarly, there were no differences in these variables by study condition in complete cases (data available on request).
Consolidated Standards of Reporting Trials (CONSORT) diagram. *As part of data processing, NORC removed 23 surveys from wave 2, and 25 from wave 4 when response patterns suggested speeding (survey duration <33.3% of overall median duration), excessive skipping of items (>50% of survey items) or ‘straight lining’ responses (selecting the same response for ‘selected grid items’).
Participant characteristics by study condition
There were few differences in individual nicotine, NRT, e-cigarette or RNC cigarette false beliefs between groups at wave 4 (online supplemental table S2). Compared with control participants, NCM intervention participants, however, had lower mean false beliefs about NRT (10.9 vs 11.3) and e-cigarettes (8.9 vs 9.2), but similar mean false beliefs about nicotine (7.5 vs 7.8) and RNC cigarettes (29.8 vs 30.3) at wave 4 (table 2). Multivariable linear regression models controlling for baseline beliefs showed no effect of study condition on wave 2 false beliefs, regardless of analytical approach (ie, intent to treat, per protocol or complete case; online supplemental table S3). Multivariable models with wave 4 false beliefs, however, indicated that the NCM condition reduced false beliefs about nicotine (b=−0.28; 95% CI −0.55, –0.009), NRT (b=−0.39; 95% CI −0.74, –0.04) and e-cigarettes (b=−0.29; 95% CI −0.57, –0.01) in complete case analyses controlling for baseline beliefs.
Unadjusted wave 4 outcomes by study condition
Further, there were few effects of the NCM intervention on intention to use cigarettes, e-cigarettes, nicotine replacement products or low nicotine content cigarettes at either follow-up (online supplemental table S4). There was a positive association between NCM condition and intention to use cigarettes at wave 2 in all models, which remained in the intent-to-treat analysis at wave 4 (b=0.05; 95% CI 0.0006, 0.10). The NCM intervention also had little impact on the number of days used nicotine or tobacco products, combusted or non-combusted, in the past 30 days at wave 2 or wave 4 (online supplemental table S5). In complete case analyses, exposure to the NCM intervention reduced the number of days used non-combusted tobacco at wave 4, but this is likely to be an unreliable estimate due to the small sample size of baseline non-combusted users (n=33).
Exploratory analyses suggested that baseline tobacco use, age, sex, literacy and concern about addiction to nicotine did not moderate the effect of study condition on wave 4 false beliefs about nicotine, NRT or e-cigarettes (table 3). However, there were main and interaction effects of past 30-day tobacco use and concern about addiction to nicotine on wave 4 RNC cigarette false beliefs. Main effects supported fewer false beliefs about RNC cigarettes among those in the NCM condition (vs control; b=−0.93) and those reporting past 30-day tobacco use (vs no use; b=−1.76); people in the NCM condition who had used tobacco products, however, had higher false beliefs than would have been expected by the joint effect of those two variables (b=−1.01), suggesting attenuation of the intervention effect. Similarly, concern about addiction to nicotine moderated the relationship between NCM and RNC false beliefs, such that those in the NCM condition who were concerned about nicotine at baseline had higher false beliefs about RNC cigarettes than would have been expected by the combination of study condition and concern at baseline (b=−0.78).
Adjusted linear regression models examining potential moderators of the relationship between study condition and false beliefs at wave 4 (n=609; per protocol analysis)†
Examination of wave 4 outcomes by dose in per protocol analyses suggested that compared with the control condition, exposure to all three NCM doses produced consistent effects on nicotine false beliefs (b=−0.33; 95% CI −0.60, –0.07), NRT false beliefs (b=−0.49; 95% CI −0.85, –0.14), e-cigarette false beliefs (b=−0.32; 95% CI −0.59, –0.05) and RNC cigarette false beliefs (b=−0.64; 95% CI −1.26, –0.02), controlling for baseline beliefs (figure 2).
Adjusted estimates and 95% CIs of nicotine, nicotine replacement therapy (NRT), e-cigarette and reduced nicotine content (RNC) cigarette false beliefs at wave 4, by dose of exposure to nicotine corrective messaging (NCM). Note: All models control for baseline false beliefs.
Discussion
In a study of 794 US adults, three exposures to eight NCMs over 3 months reduced false beliefs about nicotine, NRT, e-cigarettes and RNC cigarettes at follow-up. The emergence of NCM effects on all four products following sufficient dose is promising and underscores the importance of conducting well-controlled studies with high ecological validity to inform public education on nicotine. While these results were consistent with our hypothesis, the magnitude of the effects was smaller than in our pilot study10 and was only seen for all nicotine products after multiple exposures to intervention messages. Additionally, past 30-day tobacco use at baseline attenuated intervention effects related to RNC false cigarette beliefs at follow-up as hypothesised but, counter to our hypothesis, did not impact nicotine, NRT or e-cigarette false beliefs. A novel finding from our study was that concern about addiction to nicotine also attenuated the relationship between the intervention and RNC false beliefs, producing smaller effects than expected; these results suggest that concern about addiction to nicotine may induce resistance to NCM. There were few effects of the intervention on intention or use of nicotine and tobacco products, which may be due to the short duration of the study.
Studies over the past 20 years have repeatedly shown that smokers in the USA and abroad have little knowledge of NRT and equate the harms of NRT use with the harms of cigarette smoking.5 27 28 36–40 Nicotine misperceptions have been identified as a barrier to NRT use among smokers39; they may also limit uptake of FDA-authorised MRTPs. In line with expert consensus,41 more than 50% of adult smokers in 2007 who held incorrect beliefs about NRT reported that being exposed to scientific information to correct their misperceptions would make them more likely to use it in a quit attempt.36 Our intervention reduced false beliefs about NRT and e-cigarettes regardless of dose, highlighting these products as potential targets for brief interventions that could facilitate product switching among adults who smoke cigarettes. However, these findings also suggest resistance to changing beliefs about combusted tobacco products, which are overwhelmingly responsible for the death and disease associated with tobacco use.20
Study findings regarding RNC cigarettes are particularly important given the introduction of these products to the commercial market after completion of our trial. In December 2021, FDA authorised the RNC cigarettes VLN King and VLN Menthol King as MRTPs,42 representing one-third of the products designated with this status.43 MRTP authorisation of these products and impending FDA rulemaking on nicotine content in cigarettes and certain other combusted products are likely to rapidly increase the number of RNC cigarettes and cigars in the tobacco market. The fact that the NCM intervention impacted false beliefs about RNC cigarettes only among those who received all message exposures aligns with findings of previous studies10 11 15 16 demonstrating the difficulty in communicating nuance about risks of nicotine and combusted tobacco use to encourage proper understanding of RNC cigarette risks. Our study underscores the need for repeated exposures to nuanced messages on nicotine that address nicotine more comprehensively and in NRT, e-cigarettes and RNC cigarettes.
Strengths of our study include the general population sample of US adults, randomised controlled design, use of a message series rather than individual messages, multiple follow-ups and assessment of message impacts on intentions and use of tobacco products. Further, we did not assess message response following intervention exposures as in our pilot study,10 providing a more valid estimate of message effects as delivered in a real-world setting. Limitations of our study include the lack of qualitative data from participants to better understand the impact of the intervention on study outcomes, limited visuals in the messages and the lack of cognitive testing of individual study messages to improve the effectiveness of future NCMs.
Conclusion
Our trial in a large population sample of US adults moves beyond trials of a single exposure to a single message or a single product to understand how a more comprehensive nicotine public education campaign could reduce false beliefs about nicotine, NRT, e-cigarettes and RNC cigarettes. Findings suggest that multiple exposures to NCM over 3 months, the Centers for Disease Control and Prevention’s recommended duration for a widespread public education effort like a mass media campaign,44 would reduce false beliefs about nicotine, NRT, e-cigarettes and RNC cigarettes. Weakened effects of NCM in adults concerned about nicotine addiction suggest that the effectiveness of adult education campaigns could be hampered by nicotine and tobacco prevention messaging that increases fear of nicotine addiction.45 Future studies will improve our understanding of the dose and duration of nicotine education needed to shift intentions and behaviour related to nicotine and tobacco product use, as well as a tailored approach for tobacco product users to achieve similar reductions in false beliefs as non-users.
Data availability statement
Data are available upon reasonable request. Data are available upon reasonable request. The corresponding author will make deidentified participant data and the data dictionary available following publication. Institutions and individuals wishing to access any resources or data must contact the corresponding author (andrea.villanti@rutgers.edu). Data will only be made available to those whose proposed use of the data has been approved by the corresponding author. The recipient must agree to not transfer the data to other users and that the data are only to be used for research purposes. The investigators will require requestors of data to sign a data sharing agreement that will ensure (1) use of the data is only for research purposes, (2) data security using appropriate technology/firewalls, (3) destruction of data after data analysis and (4) proper citation in publications or other written materials. A record of transfer of data and a copy of the dataset that was distributed will be kept by Rutgers University.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants but this trial was deemed not human subjects research by the University of Vermont Institutional Review Board. Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors ACV and AAS conceived and designed the study. CP-M managed the study and was overseen by ACV. ACV conducted and is responsible for the data analysis. ACV wrote the initial draft of the manuscript. CP-M, MM, DM, ECD, JNC and AAS reviewed, edited and approved the final version. ACV had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. ACV and AAS accept full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.
Funding Research reported in this publication was supported by the National Institute on Drug Abuse of the National Institutes of Health (Award No R01DA051001).
Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Competing interests CP's spouse is employed by Perrigo, which markets consumer smoking cessation products. The other authors have no conflicts of interest to disclose.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.