Study design

A two-arm parallel-group randomised controlled feasibility trial, allocating smokers to either ‘usual care’ (link to NHS SmokeFree website)21 or the intervention arm who receive ‘usual care’ plus access to the Quit Sense app.

Additionally, there is a nested qualitative process evaluation with a subsample of control and intervention participants to (1) assess intervention experiences and how these might help explain causal pathway(s) towards smoking behaviour change (intervention arm only)22 and (2) asses experiences of study participation.

Additional information on study configuration can be found in online supplemental appendix 1.

Participants

Study setting

Participants will complete screening, consent, baseline and follow-up measures online. This online design matches the ‘real world’ setting of smartphone app uptake.

Online recruitment

Recruitment is through paid-for online advertisements with Google Search and social media (Facebook and Instagram), limited to England-based IP addresses and targeted at Android devices. Advertisements are managed by ‘Nativve’,23 specialists in digital marketing for research study recruitment. To maximise participation rate efficiency and in accordance with patient and public involvement (PPI) recommendations, different adverts will be tested using A/B experiments while recruitment is live. People clicking on adverts will be directed to the study website and provided with study information including a downloadable ‘Participant Information Sheet’. The screening and consent procedures are online with an e-signature confirming consent (table 1 and online supplemental appendix 2). After completing an online baseline questionnaire, participants will be randomly allocated to usual care or usual care plus Quit Sense. As described in figure 1 and informed by a prior text message smoking cessation intervention,24 we estimate that ~6625 unique visitors are required to achieve a sample size of 200.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow diagram showing study design, measures and sample sizes.

The distribution of socioeconomic status (SES) during recruitment will be monitored. Low SES is defined as individuals with a semiroutine or routine and manual occupation, class 5 in the National Statistics Socio-Economic Classification (NS-SEC),25 or who have never worked or are long-term unemployed. Targeted advertising will be used to increase low SES representation if less than 45% of our sample are categorised as low SES when 35% of the target sample is reached.

Procedure

Both arms will automatically receive a link to NHS SmokeFree via SMS after enrolment. Only the Quit Sense arm will receive an SMS with an app link and activation code. Use of the activation code will be monitored. Participants who do not initiate the app will be sent an SMS reminder 3 days postenrolment and then 5 days postenrolment asking to identify a reason for not initiating the app from a predefined list.

To attempt to increase response rates, all participants will receive an SMS prior to receiving links to follow-up questionnaires. Those who do not respond to the SMS with a survey link will be telephoned. Once the 6-week questionnaire is completed, a subsample of participants from both arms (approximately 3:1 ratio between Quit Sense and usual care) will be invited to a telephone interview to gather data on their experiences of the app (intervention only) and of participating in the study. Where practicable, we will use purposive sampling to interview a mixture of high and low SES abstainers and smokers, and for Quit Sense participants based on varying rates of app engagement. We anticipate interviewing 20 participants, depending on when data saturation is reached.

Six months and 2 weeks after enrolling (allowing for a 2-week quitting preparation period for Quit Sense participants; henceforth, referred to as the ‘6-month’ follow-up), all participants will be asked via SMS to complete the final questionnaire. Participants reporting abstinence in the previous 7 days at follow-up will be posted a saliva sample kit to be returned to ABS laboratories, UK.

Participants will receive vouchers on completion of the 6-week (£5) and 6-month (£10) surveys, return of the saliva sample (£5) and for participating in an interview (£20).

As suggested by our PPI panel, all participants will receive a postcard and SMS to thank them and emphasise their importance to the study.

Criteria for discontinuing or modifying allocated interventions

Unless a participant withdraws from the study, they will be followed-up for data collection regardless of their level of engagement or smoking status. Participants, who discontinue using the app for any reason other than withdrawal, will remain in the trial for the purpose of follow-up and data analysis.

Interventions

Quit sense APP arm

Quit Sense is a context-aware smartphone app designed for smokers wanting to quit. It is informed by learning theory (LT)26 27 and two theory-guided SMS text message systems,17 28 which are in turn informed by social cognitive theory (SCT).29 Quit Sense targets three key determinants that map onto LT and six mapping onto SCT: learnt associations between smoking and an individual’s physical environment (LT); learnt associations between smoking and mood triggered by the environment (LT); the presence of other smokers (LT, SCT); awareness of smoking triggers (antecedents) (SCT); outcome expectancies of a lapse (SCT); goal/intention for complete abstinence (SCT); self-efficacy in resisting urges to smoke (SCT); knowledge of and self-efficacy for lapse prevention strategy use (SCT). Twenty-one corresponding behaviour change techniques (BCTs)30 are used to target these determinants.

The main feature is ‘Geofence-Triggered Support’ (GTS), which is orientated around three app stages:

• Stage 1 (training): the user sets a quit date (default suggestion 7 days) and trains the app to learn about their smoking behaviour. This requires the user to report in the app each smoking episode and the situational context in real time (stress, mood, urge strength, setting description (home, working, home-working, socialising, other), presence of other smokers), while the app records geolocation. If a user reports smoking more than once in the same location (as defined by the app), a geofence (a circular virtual perimeter) is created representing a high-risk area. Location monitoring precision varies depending on the availability of GPS/WiFi/network (range ~5–50 m).

• Stage 2 (28-day abstinence challenge): postquit date, the app monitors the user’s location. A support message is triggered a minimum of 1 min after the user enters a geofence zone (typical range 1–15 min, depending on operating system). Messages are individually tailored using stage 1 context information and using any stage 2 smoking reports, providing support to help users avoid or cope with location specific triggers. Further messages are triggered after each 3-hour interval of remaining in that location (default between 8 am and 9.30 pm or as defined by the user). This also applies to ‘home-working’ messages. However, if the user has not reported ‘home-working’ within the time window (2–3 hours) in which the message is to be delivered, a ‘home’ message will be delivered instead. This is because ‘time’ of reporting is an additional requirement (to ‘location’) for home-working messages to be triggered.

• Stage 3 (maintaining abstinence): the app continues to deliver GTS for two further months but reduces the frequency by one-half every month. After 3 months postquit date, the GTS support stops, unless the user opts to switch it back on or restart their quit attempt.

Quit Sense has six additional features (relevant to all stages, unless specified otherwise):

1. Feedback is provided after each smoking report. In stage 1 this is tailored to nine characteristics collected during app initiation. Messages focus on enhancing quitting preparation, motivation and self-efficacy. In stage 2, feedback is non-tailored and orientated around lapse and relapse prevention.

2. An ‘End of Day’ (EoD) survey that users are prompted to complete, recording cigarettes smoked that day, cravings31 and self-efficacy28 29 (stages 1 and 2 only). Feedback is provided after each EoD survey, including encouragement to log smoking behaviour if there is a discrepancy between the survey and real time reports.

3. A ‘my profile’ screen including number of days quit, money saved (based on smoking rate and reported cost), a calendar showing a heat map and ‘smileys’ (use of yellow, grey and blue to denote values with corresponding happy, neutral and sad smileys) for smoking, cravings and self-efficacy for each EoD survey completed. In addition, smoking pattern feedback is provided using graphical and written summaries for smoking triggers based on the user’s historical reporting.

4. A ‘library’ of quitting advice split into six categories: ‘getting ready’, ‘boost your motivation’, effects of smoking and quitting’, ‘early days of quitting’, ‘staying quit’, ‘smoking after your quit date’. Users can write and submit their own support messages.

5. Scheduled non-tailored daily support messages delivered each morning orientated around the quit date (stages 1 and 2 only)—targeting outcome expectancies, motivation, preparation, self-efficacy (three sets of different messages to prevent message duplication in successive attempts).

6. The option for the user to reset their quit date in stage 1, if they are not feeling ready, or stage 2 if they relapse. Relapse is defined as multiple smoking episodes that are reported over 2 days during a quit attempt (either through smoking episode logs or the EoD survey). If this occurs, users are invited to either recommit and continue or reset their quit date and return to Stage 1.

Qualitative process evaluation

The aims of the qualitative process evaluation are:

1. To gather user experiences of app usage in order to inform app optimisation.

2. To understand participant experiences of participation and the unique contexts in which participants are attempting to quit.

Interview questions will probe for in-depth descriptions of intervention participants’ experiences of app usage including which types of support they liked most and least, the timing of support in relation to experienced need for support, and views on the features used. Control participants will be asked about their experience of trial participation. Both groups will be asked about their use of or interest in other smoking cessation aids and support and contextual details such as mandated movement restrictions and family setting. In addition, the Quit Sense app includes an in-built audio record feature so that users can leave feedback. The study will therefore also assess the feasibility of a user-initiated process evaluation approach where data are collected with high ecologically validity. Any relevant data submitted will be analysed alongside interview data, potentially providing a broader range of views from a more user-controlled perspective. Transcriptions from interviews or app recordings will be anonymised. Recordings will be deleted after transcription, or if not fully transcribed, after audio coding for analysis.

Database

Study data are collected and managed using REDcap (Research Electronic Data Capture) tools hosted at Norwich Clinical Trials Unit. REDcap is a secure, web-based software platform supporting data capture for research studies.32 33 Further details on data management can be found in the online Supplementary File.

Outcomes

Smoking and related outcomes

Smoking and related outcomes will provide preliminary information about the intervention’s impact. The main abstinence outcome for which we will monitor completeness of ascertainment is based on the Russell standard.36 Abstinence will be defined as self-reported abstinence in the previous 6 months allowing for no more than five cigarettes and not smoking in the previous week, biochemically validated by a saliva cotinine concentration of less than 10 ng/mL36 37 and for those using nicotine substitution (eg, Nicotine Replacement Therapy, or e-cigarettes) an anabasine concentration of less than 0.2 ng/mL, based on feedback from our testing laboratory.37 38 Thresholds will be reviewed prior to analysis in case of changes in guidance or relevant evidence. Given recent evidence that some e-liquid, believed to be that made outside of the UK, can contain anabasine,37 we will monitor the proportion of e-cigarette users who report abstinence from smoking but have an anabasine level above the chosen threshold.

Smoking cessation experts highlight the value of additional smoking outcomes based on different time periods of abstinence and time-points.39 We will therefore measure 7-day point prevalence abstinence at 6 months follow-up (self-report and biochemically verified) and 7-day point prevalence abstinence at 6 weeks follow-up (self-report).

We will collect the following hypothesised mechanisms of action at 6-week follow-up:

• Lapse incidence (any smoking, even a puff) since the initial quit attempt (if made),6 which is highly predictive of relapse.7–9

• Lapse prevention strategy use,14 which is associated with lapse prevention.10

• Self-efficacy,28 29 which prospectively predicts lapse and relapse to smoking.40

• The strength of urges to smoke measure,31 which prospectively predicts abstinence and is superior to other urge-measures in doing so,41 and frequency of urges to smoke.41

• Automaticity and associative processes subscales from the Wisconsin Inventory of Smoking Dependence Motives (WISDM-37).42

Sample size

In line with recommendations that feasibility studies with binary outcome measures recruit at least 60 participants per group (minimum of 120) and a maximum of 100 per group, the proposed sample size for this trial is 200 smokers (100 per arm).43 Participants will be randomised to usual care or Quit Sense arms on a 1:1 ratio. This sample size will enable key parameters to inform a future definitive trial to be estimable within the following precision (defined as the 95% CI half-width):

• Primary outcome completion: we estimate a follow-up rate for self-reported smoking status at 6 months of 80%, with precision of ±6%.44–46 For biochemical validation, we estimate that 75% of participants reporting abstinence will return a saliva sample by post, with precision of ±22%.28 45

• Cessation rate in usual care arm: we estimate an abstinence rate of 5% at 6 months follow-up, providing precision of±4%.47

• App installation and initiation: we estimate that 85% of intervention participants will instal the app, with precision of±7%.48

• App engagement—in our acceptability study, 71% used the app for more than 1 week (a timeframe deemed meaningful by our PPI panel). With a similar rate in the trial, precision would be ±9%. This is in line with other studies.48

Randomisation

Data analysis

The feasibility and acceptability of all measures will be assessed by their level of completeness and via interviews and app audio data.

Outcomes

Completeness of the anticipated primary outcome for a future trial (outcome i) and the abstinence rate using this anticipated primary outcome for the usual care arm (outcome ii) will be described as proportions with 95% CIs, and translated into interpretable probabilities using the Bayesian approach relevant in preliminary trials with the objective of powering the definitive trial.

Cost per recruit (outcome iii), rates of app installation (outcome iv) and completion of smoking cessation-related resource use and EQ-5D-5L (outcome v) will be described using summary statistics with 95% CIs.

We will estimate the preliminary intervention effect on abstinence, lapse incidence and use of lapse prevention strategies, using multiple logistic regression, providing ORs with 95% CIs, while adjusting for any potential baseline confounders defined by the statistical analysis plan (SAP). For the estimated intervention effect on abstinence, we will assume missing=smoking.36 To estimate intervention effects on self-efficacy, urges and smoking dependence motives, we will use analysis of covariance models involving adjustment for the baseline score and for predefined potential confounders, primarily using complete cases only. Missing outcome assumptions will be assessed through sensitivity analyses which will be defined in the SAP.

Economic evaluation

We aim to collect data (baseline and 6-months follow-up) that will enable us to inform the decision as to how cost (based on resource use data) and benefit (in terms of QoL) data are estimated in any future more definitive study. Completion rates will thereby be estimated.

In terms of costs, we will estimate those associated with the intervention (eg, app maintenance) and smoking related costs to both the individual (eg, nicotine replacement therapy) and the NHS (eg, NHS stop smoking services/GP visits). In line with National Institute for Health and Care Excellence (NICE) guidance,51 QoL will be measured via the EQ-5D-5L35 as this enables the calculation of QALY (Quality Adjusted Life Year) scores. Appropriate unit costs52 will also be attached to all items of resource-use in order to identify major cost drivers.

Based on the intention to treat approach, a preliminary cost-effectiveness analysis will be performed. In the base-case analysis, costs will be estimated from the viewpoint of the NHS and personal social services.51 We will estimate the mean incremental cost and mean QALY gain associated with Quit Sense compared with usual care. Assuming dominance does not occur (where one intervention is both more costly and less effective), then the incremental cost effectiveness ratio (incremental cost/QALY gain) will be estimated and assessed in relation to a range of cost-effectiveness thresholds (eg, £20 000 to £30 000 per QALY),51 in order to provide a preliminary assessment of whether Quit Sense constitutes value for money. The associated level of uncertainty will also be estimated.

Qualitative process evaluation data collection and analysis

Interviews will be audio-recorded and transcribed intelligent verbatim. App audio-recordings may be transcribed depending on quality and resources. If transcription is unfeasible, coding of audio files will be completed.

We will undertake an inductive thematic analysis of the first 3–4 interview transcripts,53 develop a coding framework from this and then code remaining data. We will check coding consistency through the independent coding of 10% of the data, with a further 10% if consistency is unsatisfactory. We will continue refining until the themes are agreed.

Analysis of process evaluation data will focus on identifying key themes associated with using the app and views and experiences of using other relevant apps. A second stage of analysis will work with descriptions of experiences of lapse or lapse avoidance in the context of the app and considering unique participant contexts. These will be reported as vignettes and used to explore participant experiences and to identify mechanisms of action and casual pathways towards behaviour change.

Patient and public involvement

We have a PPI panel of four former smokers who have contributed to study design including the recruitment strategy, wording of SMS messages, study information, the qualitative evaluation, the design of the study logo and the function and features of the Quit Sense app and support it delivers. The PPI panel is represented on the Trial Steering Committee.