The smoking cessation in pregnancy incentives trial (CPIT): study protocol for a phase III randomised controlled trial

UK pregnancy smoking rates remain high. One in four women smoke for part of their pregnancy and one in eight smoke throughout [13]. Stop Smoking Services (SSS) usually offer counselling plus free nicotine replacement therapy (NRT); however, only 10% of pregnant smokers use these services and as few as 3% stop smoking [14]. Effective approaches are limited. New interventions are needed to increase engagement with SSS, encourage uptake, support quit attempts and produce better outcomes [15].

Stop smoking support is freely available to pregnant women throughout the UK. Models of support differ, however, depending on where the women live. In general, two main types of support are offered which can be described as ‘specialist’ (just for pregnant women) or ‘generic’ (for all smokers including pregnant women). Within this framework, support offered commonly includes: individual/group support provided by specially trained advisers who may be nurses or midwives; support provided in the hospital setting, women’s homes or another mutually acceptable venue; at least one face-to-face counselling session with follow-up support, often by telephone, to 12 weeks after a quit date is set; and advice on use of NRT utilising various models of prescribing (e.g. nurse/GP prescribing/pharmacy).

The National Institute of Health and Care Excellence (NICE)—PH26 Smoking: stopping in pregnancy and after childbirth [15]—published comprehensive guidance in 2010 regarding services that should be provided to pregnant smokers.

The rationale for incentives is that they can stimulate behaviour change through providing an immediate reward for changes in health behaviours (e.g. smoking cessation), which is likely to be more motivating to people than more distal rewards such as health improvements. For smokers who quit, the saving of not buying cigarettes provides continued ‘value’ long after incentives have stopped. Despite the unborn child having no choice regarding tobacco exposure, the ‘extra’ cost of incentives is a deterrent for policy-makers and planners and is linked to a societal moral judgement of ‘rewarding bad habits’ [16, 17]. However, public opinion towards financial incentives is mixed, and public acceptability increases with effectiveness [17, 18]. This study can justify the use of financial incentives by providing evidence to show whether this upstream preventive intervention [19, 20] can be cost-effective and much cheaper than trying to cure smoking-related conditions downstream.

Published research using financial incentives for smoking cessation during pregnancy is limited to single-centre trials; however, as reported in two recent Cochrane reviews [21, 22], together they add up to a body of work indicating a beneficial effect that is likely to be cost-effective [23]. Combining data from nine trials of 2273 pregnant women, the 2019 review by Notley et al. [21] concluded that there is moderately certain evidence that women in the incentives groups were more likely to stop smoking than those in the control groups, both at the end of the pregnancy and after the birth of the baby—the RR at longest follow-up (up to 24 weeks post-partum) was 2.38 (95% CI 1.54 to 3.69; N = 2273; I² = 41%), in favour of incentives. The 2017 review by Chamberlain et al. [22] reported that high-quality evidence suggests that incentive-based interventions are effective when compared with an alternative (non-contingent incentive) intervention (four studies; RR 2.36, 95% CI 1.36 to 4.09). Pooled effects were not calculable, however, for comparisons with usual care or less intensive interventions (substantial heterogeneity, I² = 93%).

This body of research is still not sufficient to overcome concerns put forward by policy-makers regarding financial incentive payments [17] or to fully answer the first research question put forward by the NICE [15]: ‘Within a UK context, are incentives an acceptable, effective and cost-effective way to help women who smoke to quit the habit when they are pregnant or after they have recently given birth? Compared with current services, do they attract more women who smoke, do they lead to more of them completing the stop-smoking programme and do more of them quit for good? What level and type of incentive works best and are there any unintended consequences?’

To start to address these research questions in a UK context, our previous large (n = 612) single-centre feasibility trial in Glasgow, UK [24] added financial incentives to usual SSS care and compared outcomes with usual care alone. Smokers routinely identified at first maternity care visit were individually randomised to receive either usual SSS support only or the same support with the offer of financial voucher incentives. The first three vouchers were contingent on engagement with SSS. The last voucher (£200) could be earned by stopping without SSS support. Twenty-three per cent quit with the offer of usual care plus incentives (up to £400) and 9% with the offer of usual care alone (p < 0.001). A novel embedded health economic evaluation indicated that the intervention was highly cost-effective [23].

The context within which incentives are offered is important. Socio-demographic, geographic and organisation differences may affect future transferability of the intervention and the potential to implement a sustainable intervention in the long term [25, 26]. Adding incentives to a range of SSS models in different areas of the UK serving varied population groups needs to be tested before clear recommendations can be made.

In addition, further evidence is required to inform the cost-effectiveness debate of incentive-based schemes. The economic analysis from our feasibility trial [23] indicated relapse post-partum was the biggest area of uncertainty. Six months is the recommended period to measure long-term abstinence [27], as those abstinent at this time point tend to remain smoke-free in the long term [28].

A pivotal phase III multi-centre UK trial that includes cessation outcomes to 6 months after birth is therefore required to be able to recommend changes in policy and practice [15], and thus for SSS funders (such as the NHS or local government in the UK) to consider this approach to smoking cessation as part of mainstream services.

The proposed study will assess whether promising feasibility trial findings [24] can be transferred to other UK sites with different SSS configurations and population groups. If found to be effective and cost-effective in this multi-site trial, the simple novel ‘bolt-on’ nature of the intervention will make the trial results generalisable to a wide range of SSS and population groups, and will allow easier transfer of the intervention to other SSS within the UK and other parts of the world.

This RCT will examine, within a range of usual care pathways, the effectiveness and cost-effectiveness of financial voucher incentives when offered in addition to usual SSS support, to encourage women to attend SSS and set a quit date, to quit smoking and be abstinent towards the end of pregnancy and at 6 months after birth.

The primary objective is to determine whether the offer of financial voucher incentives in addition to usual SSS support leads to a doubling of the smoking cessation rate by the end of pregnancy.

Secondary objectives are as follows:

This study is a pragmatic, 42-month, multi-centre, parallel-group, single-blinded, individually randomised controlled superiority trial with 1:1 allocation designed to assess whether the addition of financial incentives to usual SSS helps pregnant women to stop smoking. In addition, an economic evaluation from a UK NHS perspective will assess cost-effectiveness of offering financial incentives added to usual SSS; a mixed-methods theory-driven [29, 30] process evaluation will examine barriers and facilitators to trial enrolment and future implementation of incentives in a range of contexts; and data from the feasibility trial centre in Glasgow [24], a deprived inner city, will also be analysed in an a priori meta-analysis.

An overview of the trial design is illustrated in Fig. 1.

Women will be recruited from SSS serving maternity hospitals in three of the four UK nations—Scotland, England and Northern Ireland. Participating sites include a deprived city, a deprived post-industrial suburban and rural area, a provincial city, two provincial towns, a deprived coastal city and a rural area. Each of these sites have different SSS configurations offering their own care pathway within the framework of the UK NICE guidance [15]. These include NHS/local authority-run services, generic/specialist pregnancy services, midwifery/SSS advisor-led services and opt-in/opt-out services, and represent most UK usual care pathways for smoking cessation in pregnancy. Each of the sites have between 1000 and 6000 deliveries per annum. The diversity of sites thus incorporates organisational differences and facilitates recruitment of a mix of women from different geographic and socio-economic backgrounds.

Eligible women for the trial are those who: are aged 16 years or over; are pregnant for less than 24 weeks gestation at maternity booking, or if not yet had their first antenatal appointment, less than 24 weeks gestation at time of consent; self-report as current smokers (at least one cigarette in the last week); live in the catchment area of the participating NHS site; and are able to understand and speak English in order to provide verbal telephone consent and follow-up smoking status.

Control group women will receive the offer of usual, local SSS support.

Intervention group women will receive the same offer of usual, local SSS support. In addition, they will be offered financial incentives up to £400 to engage with local SSS and set a quit date, and remain abstinent at each follow-up point throughout pregnancy. The incentives will be in the form of Love2Shop gift cards that can be redeemed in a wide variety of UK shops, none of which currently sell cigarettes. The incentive rewards structure is shown in Fig. 2.

The sample size for this phase III trial is 940 pregnant smokers. This was calculated on the basis of the primary outcome. A total of 940 participants (470 in each group) will detect a clinically significant doubling of the cotinine-validated quit rate from 7% with usual care alone to at least 14% with usual care plus the offer of financial voucher incentives, with 90% power at the 5% significance level allowing 15% loss to follow-up.

Eligible pregnant smokers will be enrolled over a 26-month period from February 2018 to March 2020. Recruiting for 18 months would have allowed all participants to be followed up to the secondary outcome point 6 months after birth. Recruiting for 26 months (including a 3-month funded extension from CRUK—see Funding for confirmation) has allowed an additional 8 months with recruitment slower than expected whilst allowing the first 75% of participants recruited to be followed up to the secondary outcome point 6 months after birth. This compromised 6-month post-partum follow-up was agreed with the funders, the ethics committee and the sponsor prior to the study start in September 2017.

Enrolment and randomisation will be performed over the telephone by GCP-trained call centre staff at the Database Management Company (Trial Contact Centre (TCC)) once the women’s contact details and eligibility data have been submitted to the secure online trial database by research staff. All calls will be audio-recorded and information obtained during the call entered directly into the database. After obtaining informed consent and baseline data, TCC staff will then press the on-screen button to randomise women and inform them of their group allocation. TCC staff will not be able to influence or predict the random allocation which is integrated into the database.

The random allocation sequence will be generated by York Trials Unit. Women will be allocated 1:1 to either the intervention or the control group using randomly varying permuted block sizes with no stratification factors. In addition, a random date between 34 and 38 weeks gestation for each pregnancy will be generated as the date for primary outcome data collection. This date will be concealed from both the TCC staff and the women.

It will not be possible to blind women or research nurses to group allocation. The TCC staff responsible for ascertaining the primary outcome measure of self-reported smoking status in late pregnancy (corroborated by saliva cotinine measurement collected by a research nurse) will, however, be blind to allocation. Women will be asked not to disclose their group status during the follow-up telephone call with the TCC. The statistician conducting analyses will have no contact with women but will not be blind to treatment allocation.

The trial consists of an intervention phase between 6 and 38 weeks gestation with five assessment points and follow-up to 6 months post-partum. The total study period for each participant will be 42–62 weeks dependent on gestation at enrolment and timing of primary outcome assessment in late pregnancy (randomised between 34 and 38 weeks gestation). See Fig. 1 for an overview of the study design and measurement time points, and Fig. 3 for the schedule of assessment and data collection.

The data management process will be run by York Trials Unit. The protocol was built on the platform from the phase II trial [34]. This has been improved and updated by York Trials Unit in conjunction with the central trial team (DMT, LS and MM) and has been used for submissions for regulatory approval.

The database is a modified version of that used in CPIT II. York Trials Unit, central trial management in Glasgow and research staff at one of the recruiting sites have contributed to the design of the modified version.

Data entry will be completed by trained research staff at local sites.

Statistical analysis will be conducted by York Trials Unit (AM and AK). All analyses will be carried out using the intention-to-treat principle unless stated otherwise. Treatment effect estimates will be presented along with the corresponding 95% confidence interval, and statistical tests will be two-sided at the 5% level, unless otherwise stated.

An economic evaluation to assess cost-effectiveness of offering financial incentives in addition to routine SSS will be undertaken from an NHS perspective. Details are the subject of an additional protocol paper to be published separately.

A process evaluation using a mixed-methods case-study approach will explore recruitment and assess ‘intervention context fit’. This is essential to understand both how the trial functions within different SSS and how applicable and generalisable the findings may be in terms of future implementation. Full details of the process evaluation design and methods are reported in an Additional file 1.

Data monitoring will be coordinated by York Trials Unit and includes some self-monitoring at sites (see backmatter).

Serious adverse events (SAEs) that are related to the intervention will be documented. It is not anticipated that the provision of shopping vouchers to women will be associated with any related SAEs. SAEs in the feasibility study were primarily due to miscarriages that were not related to the intervention. For this reason, a separate Data Monitoring Committee will not be assembled.

Stopping the trial for reasons not related to safety such as ‘futility because the required sample size cannot be reached’ will be decided by the Trial Steering Committee.

Data cleaning will be conducted by York Trials Unit (AM) and the central trial management team in Glasgow (LS).