Figure 2 shows the flowchart of participants through different components of the study. It includes the enrollment process, followed by randomization, mailing of the medication, and follow-up surveys completed at various time points.

Interested individuals will visit the study website and read a brief description of the study purpose, its procedures, and treatments provided. Those interested in participating will read the study information and consent form and provide their consent by clicking “yes” on the study website; they may also indicate whether they would like to be contacted for future studies. Individuals are permitted to participate in the study even if they do not give consent to be contacted for future studies. After completing the consent form, participants will complete the baseline survey, which will take approximately 20 minutes. The baseline questionnaire collects information on demographics, socioeconomic factors, education, ethnicity, smoking habits (cigarettes smoked per day, duration of daily smoking, and when participants first began smoking), importance and confidence to quit, substance and alcohol usage, concurrent usage of other medication(s), and any psychiatric comorbidities (ie, depression, anxiety, and schizophrenia).

In addition, if eligible, participants will be asked if they consent to provide a saliva sample for the genetic testing component of the study. If they agree to enroll in this substudy, they will complete the Big Five Aspect Scale personality test online through the study website, which will take approximately 10-15 minutes. Big Five Aspect Scale is a self-reported public domain test, which assesses the big 5 personality traits by asking 100 questions answered on a 5-point scale ranging from “strongly disagree” to “strongly agree” [14]. Individuals will be permitted to take part in the efficacy study even if they do not consent to participation in the genetics substudy. The questionnaires used have been published previously and include the Fagerstrom test for nicotine dependence [15], the BFAs for accessing the big 5 personality traits—extraversion, neuroticism, conscientiousness, agreeableness, and openness to experience [14]—and Patient Health Questionnaire for the evaluation of depression [16]. These questionnaires have been widely used in clinical studies, involving smoking, and extensively validated [14-16].

After submitting the completed questionnaires, individuals will be notified, through an on-screen message, whether they are eligible for the study. Those who are eligible will receive an email, including their medication assignment, the study information and consent form, a standardized letter to the doctor to convey information about the study, and a standard script for the study medication they have been assigned to for their physician (or other licensed prescriber) to sign. Because there are no in-person study visits, Salivette kits will be mailed out at the time of study enrollment to verify the participants’ smoking status.

In addition, baseline saliva sample will be used to measure the nicotine metabolite ratio, which is the ratio of 3'-hydroxycotinine and cotinine (COT) concentration; this is a validated phenotypic biomarker of the CYP2A6 enzyme activity [17]. Furthermore, Salivette kits will be sent at weeks 4, 26, and 52 to measure the salivary COT level for the biochemical verification of abstinence. Those who consent to the genetics substudy will receive a second email with the genetics study information and consent form and the Oragene saliva DNA kits by mail to collect a DNA sample for the pharmacogenetics component of the study. Participants will be compensated for each DNA (if applicable) and subsequent saliva samples received.

All eligible participants will be required to visit their physician or another licensed prescriber within 5 weeks from the enrollment date to sign the prescription for their assigned medication. Reminder emails will be sent after 2 weeks of the enrollment date to make sure that the participants have booked an appointment to visit their practitioner. At the visit, participants will discuss their medical history, the medications they are on, and any other concerns they have regarding the treatment. At his or her discretion, a licensed practitioner may either sign the prescription for study medication or choose not to prescribe the assigned medication to their patient. The practitioner’s office will then fax the signed prescription to a mail-order pharmacy contracted to do the study that will then fill the prescription.

Prescriptions signed by a licensed practitioner and verified to be authentic will be filled and mailed to the participants’ addresses by the contract mail-order pharmacy. In accordance with the Ontario College of Pharmacists’ standard of practice, pharmacists will call each study participant at the time of dispensing (ie, mailing) the medication to inform participants about directions to use the medication, discuss possible allergies, concomitant medications, and offer counseling.

The following inclusion criteria will need to be met before any individual is randomly assigned to one of the 2 treatment arms: Ontario resident; having a valid email address; aged at least 19 years; current daily smoker; smoking at least 10 cigarettes per day; smoked daily for at least the past year; and have an intention to quit smoking within 30 days of receiving the medication. The following are criteria that will exclude an individual from being randomly assigned to one of the 2 treatment arms: a history of psychotic disorder (schizophrenia or bipolar disorder) or eating disorder; brain injury; seizure disorder; pregnancy, lactation, or at risk of becoming pregnant; allergy or sensitivity to bupropion or varenicline; or currently taking varenicline or Champix, bupropion or Zyban or Wellbutrin, monoamine oxidase inhibitors, thioridazine, antidepressants, or other medications containing bupropion hydrochloride.

The primary method of recruitment for this study will be by word-of-mouth through family, friends, or health care providers. The Smoking Treatment for Ontario Patients Study [12,13], also conducted by the investigators of this trial, has successfully built a collaborative network of providers from public health units, community health centers, family health teams, and community pharmacies across Ontario. This network of providers has proven to be effective in disseminating new treatment opportunities to smokers in their area. Interested individuals will self-identify and will be directed to enroll through the study website. Moreover, as part of the consenting procedure for those who participated in the Smoking Treatment for Ontario Patients Program, individuals will be asked whether they would like to be contacted about new research opportunities by CAMH. Those who choose “yes” will then be included in a database forming a smokers’ registry that can be contacted in case new research opportunities come up. This smokers’ registry will be another method of recruitment used for this study. Furthermore, the public will be informed about the study through CAMH social media accounts, Smokers’ Helpline, and Facebook advertisements.

Eligible participants will be randomly assigned to one of the 2 study arms (varenicline or bupropion). Permuted block randomization in a 1:1 ratio in blocks of 100 will be employed. The randomization process will be computerized. Owing to the large proposed sample size in this study, there will be no stratification or minimization.

Study recruitment is no longer active; however, there are still few 1-year follow-ups that are yet to be completed by participants.

Weekly motivational emails and payment for mailed-in saliva samples at the 4-week time point will be used to encourage treatment fidelity and maintain engagement in the study.

In the 2 head-to-head RCTs of varenicline versus bupropion [6,7], participants were randomized 1:1:1 to varenicline, bupropion, or placebo arms with 341-345 participants in each group. In the trial conducted by Gonzales et al [6], the 9- to 52-week CARs were 21.9% in the varenicline group versus 16.1% in the bupropion group (OR 1.49, 95% CI 0.99-2.17, P=.057). In the trial conducted by Jorenby et al [7], the 9- to 52-week CARs were 23% in the varenicline group compared with 14.6% in the bupropion group (OR 1.77, 95% CI 1.19-2.63; P=.004). Because this is an effectiveness trial, we anticipate lower quit rates overall. Indeed, in our previous feasibility study [18], 7-day point prevalence abstinence (PPA) rates after 6 months (intention-to-treat analysis) were 18.9% in those who received varenicline and 17.4% in those who received bupropion.

To be sufficiently powered to detect a significant difference in the 26-week CARs, we plan to randomize 500 subjects to each medication arm. We predict that the real-world abstinence rates will be slightly lower than those observed in clinical trials and may be similar to what we found in our nonrandomized study that this protocol is based on [19]. Therefore, we assumed that the 26-week CAR in the varenicline-treated group would be 20% compared with 15% in the bupropion-treated group. At 500 subjects per group, we will have 80% power to detect a significant difference between groups at a .05 level of significance.

This is an open-label, clinical trial; therefore, the procedures will not be blinded because participants will need to visit their physician to approve their randomized medication. Pharmacists, primary care providers and research personnel will know which medication participants have been assigned to.

All eligible participants will receive 12 weekly motivational emails for the duration of treatment. The emails include tips on several things, other than the medications, which participants can do to help them quit smoking. The contents of the emails vary from week to week (Multimedia Appendix 1).

Participants will be contacted by email at weeks 4, 8, 12, 26, and 52 to complete follow-up surveys. The follow-up surveys will collect data on changes to the participants’ smoking pattern and medication use and any possible adverse reactions experienced. During each follow-up survey, participants will be asked whether they have experienced any adverse events. If the participants at any point are experiencing intolerable adverse events, they can contact the study personnel to withdraw from the study. They will be compensated for all the saliva and genetic samples that they had submitted prior to being withdrawn. There will be no modifications to interventions even if adverse side effects are reported. Instead, participants will either be discontinuing their treatment or will be advised to speak with their physician regarding the intervention. If participants withdraw from the study, they will still be asked to complete follow-up surveys.

To make an impact on the prevalence of smoking in the overall population, it is necessary for an intervention to reach a high number of people. Efficacious pharmacotherapies are available for smoking cessation but their reach is typically limited. To the best of our knowledge, this is the first study to attempt the mass distribution of prescription medications for smoking cessation using a randomized study design. The method proposed has been previously demonstrated by the investigators of this study to be logistically feasible and effective in terms of cessation rates [18] and has provided crucial evidence for an approach that has the potential to make a significant impact on cessation rates at a population level by demonstrating a way to take full advantage of the available smoking cessation aids.

All recruitment, consenting, and data collection are Web-based. The mail-order pharmacy is less expensive than dispensing through a research pharmacy, and we do not need to pay physicians to recruit their patients into studies because subjects self-identify first and then go to their physician (in Canada, all physician visits are paid for by our universal health care system). As such, this large randomized trial can be extremely cost effective with an estimated cost of under Can $150,000. We aim to provide medication to 1000 participants; therefore, the cost per enrollment is estimated to be Can $133.33. The 52-week CAR in the varenicline-treated group is assumed to be 18% compared with 12.5% in the bupropion group. Therefore, the cost per quit is estimated to be about Can $873.36 plus Can $33.70 for a visit to the prescriber if the government’s universal health plan is billed; this is much lower compared with the economic burden of continued smoking. A report published in 2015 estimated the annual health care cost in Canada to be Can $3071 per smoker [20].