Figure 1 shows the flowchart of the trial. This trial will be a randomized, sham-controlled, open-label, multicenter trial. Table 1 shows trial centers. This trial aims to assess the efficacy of CASC in patients with nicotine dependence when added to a standard smoking cessation program. The treatment group will use the CASC smartphone app integrated with a mobile CO checker and a Web-based patient management software for 24 weeks. Table 2 shows the overall follow-up schedule. The primary outcome of this study will be the CAR from weeks 9-24. We hypothesize that the CASC treatment group will have a higher CAR than the control app group from weeks 9-24. The CAR is defined as the percentage of individuals continuously not smoking even a puff during the specified period. We will conduct this trial in compliance with the Declaration of Helsinki, Medical Device Good Clinical Practice, Ethical Guidelines for Medical and Health Research Involving Human Subjects, and all other applicable laws and guidelines in Japan. This protocol and informed consent forms were approved by the Institutional Review Board at Keio University School of Medicine and affiliated institutions. Furthermore, we will use the latest version of the approved documents. This study is registered at the University Hospital Medical Information Network Clinical Trials Registry (UMIN000031589).

We will recruit nicotine-dependent adult participants from October 2017 to September 2018 or until the recruitment process completes; we plan to follow them until October 2019. Participants who meet all inclusion criteria will be included in the trial and those fulfilling the exclusion criteria will be removed from this trial (Textbox 1). We will obtain written informed consent from all trial participants. Participants should understand the content of the consent form sufficiently before their acceptance. Consent forms must be dated and signed both by trial participants and researchers. When consent is obtained, the first copy of the form will be kept in each institution, while the other part will be kept by participants and will not be collected even after the trial is over. In addition, we will inform all participants that their medical care will not be affected if they refuse to enroll in this trial. Participants will be able to leave the trial whenever they want to discontinue it. All participants will receive a compensation of 10,000 yen per clinic visit (7 visits in total).

Randomization will be performed for eligible participants. They will be allocated 1:1 to both arms—the CASC treatment group and the control app group. Randomization will be performed by a computer-generated random sequence with stratification for smoking cessation medications.

In addition to the 12-week standard treatment procedure for smoking cessation [5], we will prescribe the CASC to the treatment group. On the other hand, a control app will be prescribed to the control group. The standard 12-week smoking cessation treatment protocol provides 5 on-site examinations and in-person counseling by a primary physician at each outpatient clinic within 3 months.

CASC consists of a CureApp Smoking Cessation smartphone app based on a cloud system paired with a mobile CO checker for participants and a Web-based patient management software for primary care physicians. Participants in the treatment group will install the app in their smartphones and then begin taking a few minutes every day for watching a video tutorial regarding smoking cessation; in addition, participants from the treatment group will keep a digital diary of quitting smoking, chat with artificial intelligence (AI) nurse, and check their exhaled CO concentration by the mobile CO checker at least once a day. Physicians will be able to follow participants’ exhaled CO concentration data and physical conditions through a secure cloud system and review them during the clinic visits (Figure 2).

On the other hand, participants from the control app group will be able to perform only 6 basic functions in their app as follows: (1) show the user guide (how to use the app); (2) enter their profile and set the start date of smoking cessation; (3) display the course of 5 visits during the 12-week treatment with a summary of objectives of each visit; (4) show the date of next appointment; (5) get the contact form for technical support; and (6) display an app version, privacy policy, and other administrative information. The other functions of the CASC app will not be incorporated in the control app. The control app will not include the mobile CO checker either. Note that exhaled CO concentrations will be measured by the medical staff at each clinic visit regardless of the study groups, and these CO data at the clinic will be used in the analysis of study outcomes.

The CASC, control app, and mobile CO checker delivered by CureApp Inc. (Tokyo, Japan) should be carefully managed to ensure they are not used outside the clinical trial at each medical institution. A prescription code will be required to activate the investigational device, and the control group app will be issued by sponsors specific to each practicing medical institution; these sponsors regularly (at least once a year) perform inventory confirming there will be no usage of the app outside the trial.

The CASC smartphone app was developed at CureApp Inc. supervised by the Division of Pulmonary Medicine, Department of Medicine, Keio University School of Medicine. The app runs on both iOS and Android smartphones. Primary physicians will provide the app prescription code to participants at their outpatient clinics. Participants will download the app through their smartphones, activate the app entering the code, and enter their baseline data including age, sex, years of smoking, number of cigarettes per day, prescribed medications (nicotine patch or varenicline), and motivation and self-confidence regarding smoking cessation. These data will be securely sent to the cloud storage, and the AI of our system will assemble tailor-made smoking cessation programs suitable to each participant based on their personal information. Personal data stored on the cloud can be reviewed by primary physicians. The CASC app has the following 4 main components to maximize the therapeutic effects of pharmacological therapy and counseling provided by health care professionals: (1) keeping a smoking cessation diary (once a day); (2) lectures and educational videos helping its users quit smoking; (3) interactive counseling by chat-bot (AI nurse); and (4) daily measurement and recording of exhaled CO concentration levels at home through a mobile CO checker.

Regarding a Web-based personal computer software for primary doctors, it provides a data management app from patients’ CASC mobile apps and advice for physicians to follow the national clinical guidelines appropriately.

The primary outcome of this study will be the biochemically validated CAR from weeks 9-24. The success of smoking cessation has been defined as (1) self-reported continuous abstinence from weeks 9-24 and (2) exhaled CO concentration no more than 10 ppm at both weeks 12 and 24. In current smoking cessation programs, the CAR from weeks 13-24 drastically decreases, whereas our investigational device is expected to prevent this reduction by covering “treatment gaps” to promote the retention of appropriate recognition and behavioral changes. A single-arm prospective pilot study showed that a CASC app, even without a mobile CO checker, had a significantly higher CAR from weeks 9-24 than the historical cohort that had received standard smoking cessation program [16,17].

In addition, we will evaluate the following secondary outcomes: (1) 7-day point prevalence abstinence at weeks 4, 8, 12, 24, and 52; (2) the CAR from weeks 9-12 and from weeks 9-52; (3) changes in several scores related to smoking cessation, including the Mood and Physical Symptoms Scale (MPSS) [18], 12-item French version of the Tobacco Craving Questionnaire translated into Japanese (FTCQ-12) [19], Kano Test for Social Nicotine Dependence (KTSND) [20], and time to first lapse after initial visit; (4) usage rates of smartphone apps; and (5) the presence of product problems or adverse events.

Table 2 shows the overall follow-up schedule of this trial. At registration, patients’ background profile, exhaled CO concentration, nicotine dependence, and smoking status will be assessed using instruments such as the Tobacco Dependence Screener [21], Brinkman index (multiplied by cigarettes per day and years of smoking), Fagerström Test for Nicotine Dependence (FTND), FTCQ-12, KTSND, and MPSS. The background profile includes age, sex, body weight, years of smoking, cigarettes per day, and medical history at the initiation of device use. Regular follow-up visits will be scheduled at 2 weeks (±7 days), 4 weeks (±7 days), 8 weeks (±14 days), 12 weeks (±14 days), 24 weeks (±28 days), and 52 weeks (±28 days); during these follow-up visits, we will examine patients’ scores regarding FTCQ-12, KTSND (after 8 weeks), and MPSS, as well as exhaled CO concentration, status of nonsmoking, status of device use, and adverse events. Adverse events and concomitant pharmacological therapy will be recorded through the trial.

An independent monitoring staff will conduct on-site data monitoring. In addition, the staff will review trial database to confirm whether principal and subinvestigators at each clinic perform the clinical trial according to the research proposal and standard operation procedures, and data queries will be asked if necessary.

Participants will be discontinued from follow-up visits for this trial when they meet either of the following criteria: (1) participants request discontinuation; (2) researchers consider discontinuation is necessary owing to severe adverse effects, for example; (3) the eligibility criteria was violated; (4) significant deviation from the protocol occurred; (5) study institution terminated trial participation because of good clinical practice violation, for example; (6) participants’ further follow-up visits are considered very difficult to accomplish; (7) continuing smoking even at the 9^th week or participant no longer intends to continue the smoking cessation program; (8) sponsor cannot continue the trial; and (9) any other cause in which researchers decide discontinuation is appropriate.

The primary endpoint will be the CAR from weeks 9-24. We estimated conservatively 52.8% in the CASC treatment group and 40.8% in the control group as the primary endpoint (CARs from weeks 9-24) based on the survey of smoking cessation [22]. Regarding the effect size, we referred data from a clinical trial of maintenance therapy with varenicline [7], which indicated that 24 weeks of varenicline therapy showed 12% of additional abstinence rate than placebo in the total population. Thus, we estimated that 24 weeks of the CASC intervention could obtain additional effect of 12% on abstinence rate compared with the control group. A sample size of 290 in each treatment group will have 80% power using a chi-square test with a 2-sided significance level of .05 (nQuery Advanced 8.2.1).

All statistical analyses will be performed according to the intention-to-treat principle and using 2-sided at a .05 significance level. The baseline characteristics will be described by the mean and SD, median and quantiles (continuous variables), or proportion (categorical variables). We will examine the outcomes using the full analysis set. For the primary outcome, we will compare CARs from weeks 9-24 comparing the CASC treatment group and the control group using a logistic regression model adjusted for prescribed smoking cessation medications. In addition, secondary outcomes at each defined period will be evaluated with logistic regression analysis adjusted for appropriate covariates between the groups. We will provide descriptive statistics, odds ratios, 95% CIs, and P values for each outcome. A statistical analysis plan detailing all statistical computations will be completed prior to the lock of the database. SAS statistical software, version 9.4 or upper version (SAS Institute Inc), will be used for all the analyses.