Smoking Cessation Intervention for severe Mental Ill Health Trial (SCIMITAR): a pilot randomised control trial of the clinical effectiveness and cost-effectiveness of a bespoke smoking cessation service

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 07/41/05. The contractual start date was in December 2010. The draft report began editorial review in December 2013 and was accepted for publication in April 2014. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

none

Copyright statement

© Queen’s Printer and Controller of HMSO 2015. This work was produced by Peckham et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Smoking and severe mental illness

People with severe mental ill health (SMI), such as schizophrenia and bipolar disorder, are more likely to smoke1 and to smoke more heavily2 than the general population. The point prevalence of smoking among those with SMI has been estimated to be between 58% and 90%. 1,3 The presence of mental ill health is associated with an elevated risk of smoking by a factor of 2.7 [95% confidence interval (CI) 2.4 to 3.2]. 4 Smokers with SMI are more nicotine dependent, more likely to become medically ill and less likely to receive help in quitting, than the general population. 5 There are several reasons why people with SMI are more likely to smoke:

• Compared with smokers without SMI, they smoke a greater number of cigarettes per day, and this is evident even before diagnosis. 6

• They smoke each cigarette more intensely, extracting more nicotine per cigarette. 2,7

• They are much less likely to receive advice to quit smoking from their general practitioner (GP)3 or mental health specialist. 8

People with SMI have a lot of time on their hands, and smoking is part of the ‘culture’ of mental health services (among both staff and patients). In addition, people with SMI often lack self-esteem and see the future as bleak; as a consequence, they may not be motivated to look after their physical health. 5 Many people with severe mental illness are also misinformed about the risks and benefits of smoking and those of nicotine dependence treatment. 9,10 They often fear and overestimate the medical risks of nicotine replacement therapy (NRT). 11 Many believe that smoking relieves depression and anxiety12 (although nicotine actually causes anxiety). Nicotine may also improve some aspects of cognitive dysfunction in schizophrenia, which could be a disincentive for patients to quit smoking. 13

Smoking contributes to the general poor physical health of those with SMI; in the UK the standardised mortality ratio (SMR) for all causes of death among people with schizophrenia has been reported to be 289 (95% CI 247 to 337), which means that people with schizophrenia have a mortality risk of just under three times that of the general population. 14 Although people with SMI are more likely than the general population to smoke, there is evidence that this is less likely to be recorded in primary care records or to be acted on for these patients than for the general population. 15 Burns and Cohen16 found that, although the annual general practice consultation rate is significantly higher among people with SMI (13–14 consultations a year) than in the general population (about three consultations per year), their health records are significantly less likely to include data relating to a variety of health promotion areas, including smoking advice. Recent studies show that people with mental health problems are just as likely to want to stop smoking as the general population – and are able to stop when offered evidence-based support. 17,18 However, research also shows that effective stop smoking treatment is not always offered to them. 19

It is within this context that a number of policy initiatives have emerged, which emphasise improving the physical care of those with SMI, including taking initiatives to facilitate smoking cessation and the promotion of smoke-free environments in secondary care services. 5,20 This has recently been the subject of National Institute for Health and Care Excellence (NICE) recommendations about the provision of smoking cessation services for people with SMI in order to help address this health inequality. 21

Existing knowledge

Smokers most commonly cite stress relief and enjoyment as their main reasons for smoking,22 although the major cause is nicotine dependence. Nicotine acts in the midbrain, creating impulses to smoke in the face of stimuli associated with smoking,23 and producing what may be thought of as a kind of ‘nicotine hunger’ (a feeling of need to smoke) when blood nicotine concentrations are depleted. Smokers also experience nicotine withdrawal symptoms: unpleasant mood swings and physical symptoms that occur on abstinence and are relieved by smoking. 24 Nicotine dependence is the main reason that most unassisted quit attempts fail within a week. 25 Cochrane systematic reviews24–34 and evidence-supported guidance from NICE,35,36 highlight that the following smoking cessation interventions (including medications used as smoking cessation aids) are helpful in helping smokers reduce their tobacco intake and quit smoking.

Rationale for the Smoking Cessation Intervention for Serious Mental Ill Health Trial

Despite the higher prevalence of smoking, a substantial proportion of people with SMI express a desire to quit. In a large population-based cohort, 30.5% of smokers with past-month mental illness had a self-reported ‘desire to quit’ (although this is lower than the rate of 42.5% among smokers without illness). 4 The introduction in 2004 of a new General Medical Services (GMS) contract46 created a policy impetus to improve the quality of primary care in priority areas. In terms of mental health, the new GMS contract specified that primary care is responsible for the provision of physical health care. Importantly, for smoking cessation initiatives, it ‘incentivises’ GPs to (1) produce a register of people with severe long-term mental health [Quality and Outcomes framework (QOF) indicator MH8 – the SMI register47] and (2) ensure that at least 90% of SMI patients have had a review that includes smoking status recorded within the previous 15 months (the QOF indicator MH9 – SMI health check48). This check includes patients seen in primary care, in secondary care and under shared care arrangements.

For those who are admitted to hospital, the introduction of smoke-free polices provides an opportunity to address smoking. This ban includes inpatient psychiatric units, although the complexities of the Mental Health Act49 have been interpreted by some hospitals as a requirement to provide smoking areas. The admission of an individual to hospital, while being stressful and occurring at a time of personal crisis, also provides a unique opportunity to provide general health advice and to engage individuals in interventions targeted at smoking reduction and cessation.

Recent guidance issued by NICE50 offers clear statements of purpose to make secondary care services (including mental health services) entirely smoke free and to promote a smoke-free culture among staff and users of services. Mental health services are highlighted as areas of priority and unmet need in relation to smoking cessation and there is clear guidance that services should be developed and implemented as a matter of some priority.

Smoking cessation services for people with SMI are not sufficiently evolved or embedded within the NHS. From the preceding discussion, we know ‘what works’ for smoking cessation in general; the purpose of the Smoking Cessation Intervention for Serious Mental Ill Health Trial (SCIMITAR) is to use enhancements of care to ensure that evidence-supported interventions are offered to (and taken up by) people with SMI and to see if smoking rates can be reduced. This technology represents a ‘complex health-care intervention’, and this study, therefore, uses the stepwise Medical Research Council complex interventions framework51 and the updated guidance52 to evaluate the clinical effectiveness, implementation and content of a bespoke smoking cessation (BSC) service for people with SMI.

Research objectives

1. To develop a BSC service based upon evidence-supported treatments for people with SMI.

2. To establish the acceptability and uptake of this BSC service by people with SMI in primary care and specialist mental health services.

3. To test the feasibility of recruitment and follow-up in a pilot trial of a BSC service among patients with SMI.

Study design

This study was a pragmatic, two-arm, parallel-group, pilot RCT. The setting was in primary care and specialist mental health services within three centres: York/Scarborough [principal investigator (PI) Professor Simon Gilbody), Manchester (Investigator Professor Helen Lester) and Hull (PI Professor Simon Gilbody). We recruited from both primary and secondary care settings. Given that this is a hard-to-reach population, several methods were used to try to identify and recruit eligible participants. A two-stage recruitment process was employed to check for eligibility, understanding of the study and to obtain consent. Participants were individually randomised to receive usual care or usual care plus a BSC service. Participants were followed up over the course 12 months, with data collected at 1 month, 6 months and 12 months post randomisation.

Approvals obtained

Ethical approval was sought and granted on 29 October 2010 by Leeds (East) Research Ethic Committee (10/H1306/72). Approval was also obtained from the relevant research and development departments (see Appendix 1).

Trial sites

The study was conducted in three sites in England. Sites recruited throughout the duration of the study.

Participant eligibility

Identifying participants

We used four methods to recruit participants: direct GP referral or following database screening, primary care referral following annual health check, secondary care recruitment – Care Programme Approach (CPA) and via Community Mental Health Teams (CMHTs) – and patient self-referral.

Screening for eligibility

Consenting participants

Potential participants who met with the SCIMITAR researcher were given the opportunity to clarify any points they did not understand and ask any questions. A full oral explanation of the trial was given by the SCIMITAR researcher. It was emphasised that the participants may withdraw their consent to participate at any time without loss of benefits to which they otherwise would be entitled. Participants were also informed that, by consenting, they agreed to their GP being informed of their participation in the trial and that their medical records may be inspected by regulatory authorities, but that their name would not be disclosed. Written informed consent was then obtained, with both the participant and the researcher signing and dating the consent forms prior to the patient being randomised.

Baseline assessment

Once participants had consented to take part in the trial, they completed the baseline questionnaires. In addition, height and weight measurements were taken in order to calculate the patient’s body mass index (BMI) and an exhaled breath carbon monoxide (CO) reading was taken. These made up the participants’ baseline data set. The participant was randomised on completion of this data set.

Randomisation

Simple randomisation was used following a computer-generated random number sequence. The SCIMITAR researcher contacted a secure randomisation line run by the York Trials Unit and, once given the details of the patient’s allocation, immediately informed the patient of his or her allocation and set up the first appointment with the MHSCP (if so allocated). A letter was sent to the GP and mental health specialist to be included in the patient’s records and to advise them on subsequent smoking cessation management. Owing to the nature of the intervention, it was not possible to blind participants, GPs, researchers or the MHSCPs to the treatment allocation.

Ineligible and non-consenting participants

All ineligible and non-consenting participants were referred back to their GP practice so that any general health-care advice on the importance of stopping smoking could be provided by the patients’ GP or community nurses.

Sample size

This pilot trial aimed to recruit 100 patients with SMI to obtain preliminary estimates of the effect size of the BSC service. Using the following assumptions (1) primary care QOF registers were assumed to give prevalence data for SMI of 0.5%; (2) an average of 2.5 whole-time equivalent GPs were assumed to work in each practice each with a list size of 1600 patients (4000 per practice); and (3) at least 80% of people with SMI smoke. If we were to recruit 25% of eligible patients in primary care, around 20 practices would enable us to recruit 100 patients over a 12-month period. This was a conservative assessment which did not allow for recruitment from secondary care, where recruitment is less easy to plan but was in addition to primary care.

Description of interventions

Follow-up

Participants were followed up 1 month, 6 months and 12 months after randomisation.

Baseline assessments and 12-month follow-up were carried out face to face, while 1- and 6-month follow-ups were carried out by telephone interview, using paper questionnaires or via online questionnaires. If it was not possible to meet the participant for a face-to-face 12-month follow-up, a systematic approach was used to explore other avenues to collect data (self-report data only). Follow-up was carried out by researchers who were not blind to treatment allocation; however, the objective nature of the primary outcome eliminated any potential bias.

Outcomes

Participant engagement

Participant engagement was measured using the proportion of intervention participants who engaged with (1) contacts which were offered from a MHSCP; (2) medication when this was offered by their GP (as measured by the number of filled prescriptions issued by the GP); or (3) compliance with CO monitoring by MHSCPs. Smoking status at baseline and (where possible) follow-up were verified by exhaled CO. Readings < 10 p.p.m. confirmed that participants had not smoked recently (i.e. within 12 hours). Measurements above 10 p.p.m. indicated that the patient has not ceased smoking. At least two CO readings were taken; if participants claimed to have stopped but their CO readings were above 10 p.p.m., they were asked when they had last smoked and whether or not they had any minor relapses during their quit attempt.

Statistical analysis

All analyses were conducted on an intention-to-treat basis, including all randomised patients in the groups to which they were randomised. Analyses were conducted in STATA version 13 (StataCorp, College Station, TX, USA).

Qualitative substudy

We explored specific issues of acceptability and adherence of smoking cessation interventions among those with SMI and those who referred to and delivered the intervention. Full details of the qualitative substudy are given in Chapter 6.

Cost assessment

A cost assessment was carried out to estimate the cost of the alternative treatment strategies. A costing methodology was carried out in two steps. The first step was to measure resource use by trial patients in physical units. Health-care and community services resource use information was collected using an adapted health economic/service utilisation questionnaire included in the baseline and follow-up questionnaire (see Appendix 3). For medication use, the participating GP surgeries were asked to extract prescription information of participants during the trial period from their records at the end of the trial. Owing to the huge variety of medication that could be prescribed to participants, a list of antipsychotic medication was used to reduce the burden of data collection upon the GP surgeries (Table 2). Therefore, only the information on antipsychotic medication and prescriptions related to pharmacotherapy for stop smoking was collected from GP surgeries. The pharmacotherapy for stop smoking included NRT products, varenicline and bupropion. While the prescriptions of pharmacotherapy were collected from GP records, the usage of pharmacotherapy was also collected using the trial questionnaire, covering both prescription and over-the-counter purchases. The second step was to calculate the cost of resources used by applying market prices or national average unit costs. Costs were assessed from a NHS and Personal Social Service perspective. Intervention costs were based on delivery costs within the trial and included staff, equipment, supervision and appropriate annuitised capital costs. Missing data were estimated by using the average value among the same group at the same follow-up point.

A cost-effectiveness analysis was undertaken at 12 months comparing resource use in the BSC group with resource use in the usual-care group using the incremental quit rate for the intervention over and above usual care. This cost-effectiveness analysis is undertaken to demonstrate the analysis that would be undertaken in a full trial. This is a pilot trial and results should be interpreted with extreme caution because of the small sample size. We do not conduct a full cost–utility analysis because of the very small sample size.

Adverse events

Clear guidance on the prescription of antismoking medications in the presence of SMI (including safety considerations) have been published and were made available to all GPs to help inform their prescribing decisions. A key feature of the SCIMITAR trial was to ensure that GPs manage antismoking medications within this framework and with their prior knowledge of the patient and their concomitant use of medication. This was with the aim of replicating real-life practice of the use of antismoking medications in primary care. The medication profile of the individual participants was reviewed by their GP or mental health specialist to assess any potential safety issues (in line with the latest practice guidance on the provision of smoking cessation interventions in the NHS). An important aspect of the design of this study was that the SCIMITAR team had no direct influence over prescribing decisions made by GPs since this was not a drug trial or an investigation of a medicinal product(s).

A standard operating procedure for detecting and reporting adverse events (AEs) was implemented. An AE was defined as any unexpected effect or untoward clinical event affecting the participant. It could be directly related, possibly related or completely unrelated to the intervention. It was also classed according to severity, either as a non-serious AE (including discomfort or slight worsening of symptoms) or as a SAE (which could be particularly harmful, dangerous or required hospitalisation).

The participant’s MHSCP, GP and mental health specialist were requested to inform the research team of any serious or non-serious AEs. In addition, participant responses to questions in the follow-up questionnaire relating to hospital admissions, attendance at accident and emergency, use of the emergency services or if the participant volunteered information which could potentially be classed as an AE or SAE, were followed up by the research team.

All AEs and SAEs were independently reviewed by a clinician and reported to the Data Monitoring and Ethics Committee and Trial Steering Committee. Any SAEs that were deemed to be related to the intervention and unexpected were reported to the Research Ethics Committee and sponsor within 7 days of notification.

Suicide protocol

A protocol for identifying and reporting suicide risk was implemented (see Appendix 6 for protocol). Question 9 on the PHQ-9, which asks if the patient ‘have you had thoughts that you would be better off dead or hurting yourself in some way?’, was used to identify any suicide risk.

If the participant indicated a response of 3 for this item, then the suicide protocol was implemented and the patient asked if he or she had talked to a GP, psychiatrist or care co-ordinator/community psychiatric nurse about these feelings. If the patient had not sought help, consent was sought to to inform the patient’s GP of the situation. If the patient refused, the relevant designated psychiatrist/health professional was contacted. If the patient agreed, the patient’s GP or psychiatrist was contacted immediately. A suicidal intent form was also completed and, where applicable, a ‘suicidal intent form: psychiatrist/health professional’ was completed. These forms were stored with the patient’s trial records.

Online questionnaires

In the original proposal participants would complete paper questionnaires and the answers would be manually entered into the database by the researcher. However, the possibility of participants completing the questionnaires online became available. As some participants may find this preferable to completing a paper questionnaire, they were given the additional option of completing questionnaires online via a secure website held on the university server.

Extension to end of study date

Owing to recruitment taking longer than anticipated an extension was requested and granted. It was originally planned that the study would end in May 2013. This was extended to November 2013. The extension was to allow sufficient time for the study team to collect all outstanding follow-up data from participants.

Twelve-month follow-up

In the original protocol, all 12-month follow-ups were to be carried out face to face. It became apparent because of the nature of the patient population being studied that this would not always be possible. To collect data from as many participants as possible, we decided that if a participant could not be met face to face, attempts would be made to collect 12-month data via a telephone interview or postal questionnaire. In these cases smoking abstinence or reduction would not be verified by CO measurements; however, self-reported quit rates would still be collected.

Gift vouchers

The SCIMITAR Trial management group offered participants taking part in the qualitative substudy a £10 gift voucher as a goodwill gesture and token of thanks for their time.

Recruitment

Recruitment started in May 2011 and ended in May 2012. Over the course of the trial, 45 GP surgeries in Manchester, York/Scarborough and Hull mailed out recruitment packs. Recruitment of at least one trial participant occurred in 25 of the 45 GP surgeries which mailed out packs. Twenty-nine CMHTs were enlisted to recruit participants, along with 21 other secondary care organisations and 14 tertiary care organisations. Four participants were recruited through direct referral, having seen a poster advertising the study.

In primary care, 1036 recruitment packs were mailed out by GP surgeries which resulted in 64 consent to contact forms being returned (a response rate of 6.2%). Of these, 51 people were recruited into the study. From secondary care 57 direct referrals were received, of which 42 were recruited and randomised (a rate of 74%); however, it was not possible to determine how many packs were given out by CMHTs and other secondary and tertiary care organisations.

A total of 97 patients were recruited to the trial. The rate of recruitment is shown in Figure 1. Recruitment occurred at three sites (York/Scarborough, Manchester and Hull) and was evenly distributed between primary and secondary care (Table 3). Table 4 shows how many people were recruited from each of the different secondary care organisations. Participant flow through the trial is shown in the Consolidated Standards of Reporting Trials (CONSORT) diagram (Figure 2). The conversion rates of each stage of the recruitment process in primary care are shown in Figure 3 and the conversion rate of each stage of the secondary care recruitment process is shown in Figure 4. It can be seen that a total GP list size of 466,734 led to 51 randomisations.

FIGURE 1.

Trial recruitment rate.

FIGURE 2.

A CONSORT diagram showing participant flow through the trial.

FIGURE 3.

Primary care randomisations from GP database searches.

FIGURE 4.

Secondary care randomisations.

Of the 97 participants, 51 were randomised to usual GP care and 46 participants were randomised to BSC (Table 5).

Baseline data

The baseline characteristics of participants are summarised in Tables 6–10. Table 6 summarises participants by the prespecified prognostic factors. There were more male than female participants (59.8% vs. 40.2% respectively) and a greater proportion of men in the BSC group (69.6%) than in the usual GP care group (51.0%). There was some imbalance between the treatment groups with respect to alcohol consumption, with more alcohol consumption in the usual GP care group (62.7%) than in the BSC group (50.0%). The mean age of participants was 47 years, with a range from 19.1 to 73.3 years.

Table 7 summarises the general health of participants. Almost half of participants reported moderate health (48%) over the past year and a mean of 5.5 consultations with a GP in the last 12 months (range 0–40 consultations). The majority of participants (86%) felt that smoking had affected their health, and 63% of participants had been advised to stop smoking by their GP. The mean BMI of participants was 28.6 kg/m² (range 17.9–43.1 kg/m²), which is categorised as overweight (normal weight BMI range is 18.5–25 kg/m²). Fifteen per cent of participants reported taking recreational drugs.

Table 8 summarises the sociodemographic data of participants and Table 9 summarises their employment status. The majority of participants (87%) were white British and the next largest ethnic group was black or black British-Caribbean (5%). About 20% of the participants had General Certificates of Secondary Education (GCSEs)/O-levels as their highest educational qualification. Over half of participants (56%) were not employed but not seeking work because of ill health and 75% of those unemployed had not been employed for over 5 years. Over half of participants were single (57%), 20% were married or living with a partner and 19% were divorced or separated.

Table 10 summarises baseline mental health status. The most common severe mental health problems were schizophrenia or other psychotic illness (n = 57, 59%), schizoaffective disorder (n = 10, 10%) and bipolar disorder (n = 30, 31%). Over half of the participants (56%) had a CPA co-ordinator and 60% were under the care of a CMHT. On average, participants had twice (mean) in the last 10 years required psychiatric treatment in hospital, with a range of 1 to 15 periods of hospital treatment. Eighty per cent of participants described their condition as ‘stable’ and 8% described their condition as ‘unstable’ (though each participant had been judged to be stable from the point of view of their condition by either their GP or a responsible mental health professional). Almost all participants (99% of those who responded to the question) were taking a medication, the most common being olanzapine (23%) and clozapine (8%).

Table 11 summarises smoking history of participants. The mean length of smoking was 27 years, with a range from 3 years to 60 years. Most participants smoked packet (66%) or hand-rolled cigarettes (53%) (or both). The median number of attempts to quit was three, with a range of 0 to 150 attempts. The mean duration of reported longest quit attempt was 43 days (median 8.5 days), with a range of 0 to 832 days. The most common self-reported previous strategies used to stop smoking were ‘cold turkey’ (70%), followed by nicotine skin patches (68%), nicotine chewing gum (52%) and nicotine inhalator (47%).

The reasons for smoking and their importance are summarised in Table 12. The most important reason given for smoking was helping to cope with stress (65%), followed by helping to relax (47%).

Table 13 summarises the reasons for giving up smoking. The most important reason cited for trying to give up smoking was that it is bad for health (86%).

Table 14 summarises the smoking behaviour of participants at baseline. Most of the participants smoked more than five cigarettes in the last week (96%). The mean number of cigarettes smoked per day is 25, with a range of 5 to 60. The mean CO reading is 24 p.p.m. with a range of 4 to 58 p.p.m. A CO reading > 20 p.p.m. indicates heavy smoking. The majority of participants (80%) said they smoke the same number of cigarettes every day. Table 15 summarises recent quit attempts. The median number of quit attempts in the last 6 months was three, with a range of 0 to 150 attempts. The mean length of the most recent quit attempt was 23 days, with a range of 0 to 180 days.

Table 16 gives a breakdown of the answers to the FTND questionnaire score at baseline. The mean FTND score at baseline was 6.6, indicating moderate nicotine dependence, with a range of 0 to 10. The mean MTQ questionnaire score at baseline was 13.8, with a range of 6 to 19. Table 17 summarises the PHQ-9 and SF-12 scores. The mean PHQ-9 score at baseline was 9.2, indicating moderate levels of low mood but below the threshold for case-level depression (indicated by a score ≥ 10). The mean SF-12 physical component score at baseline was 45, with a range of 15 to 67. This is lower than the mean of the general UK population, indicating worse physical health than the general population. The mean SF-12 mental component score was 41, with a range of 13 to 64. This is about one SD lower than the mean of the general UK population, indicating worse mental health than the general population.

Withdrawals

There were 15 participant withdrawals (15%) from the trial: five (10%) from the usual GP care group and 10 (22%) from the BSC group. A total of seven participants (7.2%) withdrew fully from the trial, while five participants (5.1%) withdrew from follow-up and three participants (3.1%) were too unwell to continue (Table 18). There are four categories of patient withdrawal:

• Full withdrawal – participant withdrawn from the trial with regards completion of both postal questionnaires and collection of GP data.

• Withdrawal from follow-up – participant has withdrawn from the completion of postal questionnaires, but agrees with the continuing collection of GP data.

• Withdrawal from treatment – participant withdraws from trial intervention treatment, but agrees with continuing completion of postal questionnaires and collection of GP data.

• Too unwell to continue – participant is deemed too unwell by medical staff to complete any questionnaires. This generally only occurs when a participant has been hospitalised.

Follow-up

Participants were given the option of providing data face to face, via the telephone or by postal questionnaire. Of those who returned follow-up data at 12 months only one person declined a face-to-face visit and completed the follow-up by telephone; all the other participants who completed a 12 month follow-up did so face to face. Participants did not use the option of completing questionnaires online.

Primary outcome

Smoking cessation at 12 months was defined as a CO measure of < 10 p.p.m. or self-reported cessation if no CO measure was available. A CO measure of < 10 p.p.m. indicated no smoking in the last 8 hours and self-reported quit indicated no smoking within the last week. At 12 months, 64 participants had a CO measure and four participants had only a self-reported measure. Eight out of thirty-five participants (23%) had stopped smoking in the usual GP care arm and 12 out of 33 participants (36%) had stopped smoking in the BSC arm (Table 19).

A logistic regression of smoking cessation at 12 months on randomised group, adjusted for sex, age, number of cigarettes smoked at baseline and alcohol consumption at baseline, gave an OR of 2.9 (95% CI 0.8 to 10.5) for BSC compared with usual care (Table 20). This indicates that those randomised to BSC have greater odds of smoking cessation than those randomised to usual care, although this is not statistically significant. However, the analysis has been carried out on a small sample (complete cases, n = 65), so results should be interpreted with caution.

Secondary outcomes

A summary of the smoking-related secondary outcome is given in Table 21 and a summary of the non-smoking-related secondary outcomes is given in Table 22.

The FTND questionnaire produces a score between 0 and 10, where a score of 1–2 indicates low dependence, 3–4 indicates low to moderate dependence, 5–7 indicates moderate dependence and 8–10 indicates high dependence. The FTND score slightly decreased over time for both randomised groups (Table 21). The usual-care group has a lower FTND score than the BSC group at 12 months, but the CIs overlap (Figure 5).

FIGURE 5.

Mean FTND questionnaire score and 95% CI by randomised group.

The MTQ questionnaire produces a score between 4 and 19, where a higher score indicates greater motivation to stop smoking. The MTQ score decreases over time in both randomised groups (Figure 6). The MTQ score is higher in the BSC group than in the usual-care group at all three time points and the CIs are not overlapping at 1 month.

FIGURE 6.

Mean MTQ questionnaire score and 95% CI by randomised group.

The PHQ-9 produces a score between 0 and 27, where scores of 5, 10, 15 and 20 are used as cut-off points for mild, moderate, moderately severe and severe depression respectively. The PHQ-9 score is fairly stable over the first 6 months and then increases at 12 months for the BSC group and decreases for the usual-care group (Figure 7). The PHQ-9 score is higher (indicating lowering of mood) in the BSC group than in the usual-care group, but the CIs are overlapping.

FIGURE 7.

Mean PHQ-9 score and 95% CI by randomised group.

The SF-12 physical component score ranges from 0 to 100, where 0 indicates the lowest level of health and 100 indicates the highest level of health measured by the scale. The physical component score appears to be fairly stable over time, increasing slightly at 6 months and then decreasing at 12 months (Figure 8). The physical component score is slightly higher in the BSC group than in the usual-care group at 6 and 12 months, indicating better physical health, but there is very little difference between the two groups.

FIGURE 8.

Mean SF-12 physical component score and 95% CI by randomised group. PCS, physical component score.

The SF-12 mental component score ranges from 0 to 100, where 0 indicates the lowest level of health and 100 indicates the highest level of health measured by the scale. The mental component score appears to be fairly stable over time for the usual-care group and decreases in the BSC group (Figure 9). The mental component score is lower in the BSC group than in the usual-care group, indicating lower composite mental health, but the CIs overlap at all time points.

FIGURE 9.

Mean SF-12 mental component score and 95% CI by randomised group. MCS, mental component score.

The number of cigarettes smoked per day appears to be fairly stable over time (Figure 10). At 12 months, fewer cigarettes were smoked in the BSC arm than in the usual-care group.

FIGURE 10.

Histograms of number of cigarettes smoked per day by randomised group at 12 months. (a) BSC; and (b) usual GP care.

The number of attempts to quit in the last 6 months appears to decrease at 6 months and increase at 12 months. The number of attempts to quit is greater in the BSC arm than in the usual-care arm at all time-points, including at 12 months (Figure 11).

FIGURE 11.

Histograms of number of attempts to quit by randomised group at 12 months. (a) BSC; and (b) usual GP care.

The length of the most recent quit attempt has a large range in the BSC group at 6 and 12 months (Table 22). The length is shorter at 1 month in the BSC group than in the usual-care group, but longer at 6 and 12 months. Figure 12 give details of the percentage of self-reported smoking cessation by randomised group.

FIGURE 12.

Percentage of self-reported smoking cessation and 95% CI by randomised group.

The number of participants self-reporting cessation increased slightly over time (Table 22). The self-reported quit rate was higher in the BSC group than in the usual-care group at all time-points, but the CIs overlap. A cross-tabulation of self-reported quit with CO-measured quit showed moderate agreement between the two measures (κ = 0.48). There were no participants who had quit according to the self-report measure and not quit according to the CO measure, but there were 11 participants who had quit according to the CO measure but did not self-report quitting (Table 23). However, note that a CO measure of < 10 p.p.m. indicates no smoking in the last 8 hours and self-reported quit indicates no smoking within the last week.

At 12 months, there were two participants who reported drug use in the BSC arm (6%) and four in the usual-care arm (12%).

Adverse events

There were 21 AEs among 17 participants during the course of the trial. Of these, 11 were classed as serious. More participants in the BSC group (23.9%) than in the usual GP care group (11.8%) experienced one or more AEs. The six AEs that were definitely or probably related to the intervention were all non-serious AEs (Table 24).

Of the participants who experienced SAEs, eight were hospitalised as a result of deterioration in their mental health; one participant had surgery for an existing problem, two participants were hospitalised for illnesses unlikely to be related to the study and one participant died as a result of lung cancer.

Of the participants reporting AEs, one experienced a deterioration in their mental health that did not require hospitalisation, four participants experienced side effects of NRT products (burning mouth, feeling sleepy, headaches), two participants experienced side effects of smoking cessation medication (headaches, nightmares) and one participant had an unrelated complaint (ear infection).

Summary of findings

At 12 months, 36% of participants had stopped smoking in the BSC group, compared with 23% in the usual-care group. The adjusted OR was 2.9 (95% CI 0.8 to 10.5), indicating a greater likelihood of smoking cessation in the BSC group than the usual-care group, but the difference is not statistically significant.

In terms of smoking-related outcomes at 12 months, the BSC group generally performed better than the usual-care group. At 12 months the FTND questionnaire score, MTQ score, number of cigarettes smoked per day and number of cessation attempts were higher in the BSC group than in the usual-care group. In addition, the duration of cessation was longer in the BSC group.

In terms of mental health outcomes at 12 months, the BSC group had a slightly lower mean mental component score (a measure of psychological well-being incorporating anxiety and depression) and a slightly higher PHQ-9 score (a measure of depression). Mental health outcome – as measured by these metrics – was not different between groups at either 1 or 6 months. In terms of physical health outcomes at 12 months, the BSC group fared better than the usual-care group overall, with slightly higher mean physical component scores and slightly lower mean BMI.

Costs

Intervention costs were calculated based on training costs and the costs of delivering the intervention.

Total costs

Mean total health-care costs are presented in Table 28 for BSC and usual care. Mean total cost per participant in the BSC group was £12,674, compared with £6867 in the usual-care group. The health-care resources/community services were the main cost driver in both groups, representing the majority (94%) of the costs incurred.

TABLE 28 - Cost per participant during 12-month trial period

Cost item	BSC (n = 46) (SD)	Usual care (n = 51) (SD)
BSC intervention	£221 (£160)	£0 (£0)
Antipsychosis medicine prescription	£474 (£913)	£428 (£782)
Pharmacotherapy for stop smoking prescription	£62 (£132)	£17 (£60)
Health-care resources/community services	£11,917 (£16,601)	£6421 (£6089)
Total	£12,674 (£16,595)	£6867 (£6026)

Mean costs and their SDs were calculated. The SDs demonstrated a high variance among costs, especially the cost of health-care resources/community services. The exact range of each cost component is shown in Table 29. Because of the small sample size in both groups, low-frequency, high-tariff costs can have a significant impact upon the mean cost. One participant in the BSC group stayed in rehabilitation for 6 months immediately prior to the trial and subsequently received multiple services on a daily basis for at least another 6 months. No patients in the usual-care group were observed to have similar episodes; hence, in the small sample, the average in the BSC group may be inflated by this case. This is a common problem in the analysis of pilot trials with small populations, where small sample size, even under random allocation, can result in some baseline imbalances.

TABLE 29 - Cost range (per participant) for 12-month trial period

Cost item	BSC (n = 46)	Usual care (n = 51)
BSC intervention	£37–£824	–
Antipsychosis medicine prescription	£0–£3712	£0–£3247
Pharmacotherapy for stop smoking prescription	£0–£706	£0–£300
Health-care resources/community services	£352–£96,896	£86–£33,217
Total cost	£716–£97,232	£343–£33,217

It should also be noted that the greatest proportion of cost was accounted for by the utilisation of wider health care and social services outside the trial, which was considerably higher in the intervention group (Figure 13). The highest utilisation of health care and social services occurred in the intervention group, and the utilisation level remained higher in the intervention group when the outliers were excluded (as shown by the BSC* line in Figure 13). Although there was a consistent difference in health-care and social services usage between two groups, the limited sample size prevented us from concluding a statistically significant difference.

FIGURE 13.

Mean cost of health care and social services in the last 6 months at baseline, 6 months and 12 months. UC, usual care.

European Quality of Life-5 Dimensions

Quality-adjusted life-years (QALYs) were derived from EQ-5D data by calculating the area under the curve derived from EQ-5D at each time point. 70 EQ-5D data were collected at baseline and at the 1 month, 6 months and 12 months time points. Missing data at each point were replaced with the mean value in the relevant group. Figures 14–16 show the proportion of participants who reported a problem in each individual domain of EQ-5D (i.e. domain score of 2 or 3) over the trial period in each group. Similar patterns in mobility and usual activities were evident in both groups. However, the patterns in other domains were almost opposite between the two groups. During the 12-month trial period, participants in the BSC group gained a mean of 0.65 QALYs (95% CI 0.58 to 0.72 QALYs), while participants in the usual-care group gained, on average, 0.69 QALYs (95% CI 0.63 to 0.75 QALYs).

FIGURE 14.

Mean EQ-5D score at baseline and at 1 month, 6 months and 12 months. UC, usual care.

FIGURE 15.

Proportion of the participants who reported a problem (minor or severe) in each individual domain of EQ-5D: BSC (intervention group).

FIGURE 16.

Proportion of the participants who reported a problem (minor or severe) in each individual domain of EQ-5D: usual care (control group).

We have not undertaken a full incremental cost-effectiveness analysis, as this is a pilot trial and was not powered to detect significant differences in cost-effectiveness. The aim of this pilot is to assess the feasibility of conducting a full economic evaluation and to investigate the completeness of data. We should also note that a 12-month trial may not demonstrate the potential long-term impact on health as well as health-care and social services utilisation. It is likely that cost savings and quality-of-life gains as a result of smoking cessation would extend beyond the 12-month follow-up in this study. 71 Hence, a full-trial analysis would use longer-term modelling to project costs and outcomes beyond 12 months.

Cost-effectiveness

Total health-care and social care costs were combined with the primary outcome of the trial to estimate the cost per quitter at 12 months. As this is a pilot trial, which is not powered to estimate cost-effectiveness, we report a simple incremental cost-effectiveness ratio (ICER) by combining the costs with the number of successful quitters at 12 months; the incremental cost was £58,197 per quitter. This ICER should be treated with caution because of the small sample size and large variance of total cost. In a pilot trial, high-tariff, low-frequency costs can have a large impact on the overall ICER if these high-cost cases fall into a treatment arm by chance. Therefore, the main aim of the economic analysis of this trial has been to pilot questionnaires and assess the feasibility of collecting such data in a larger trial.

Smoking cessation help beyond the trial

Beyond our analysis perspective, 19 participants in the BSC group and 14 participants in the usual-care group used resources other than NHS-funded resources regarding smoking cessation, including other helplines, the internet and self-help booklets. Among these participants, eight in the BSC group used other helplines, ranging from one to nine times during the trial period, while six participants in the usual-care group used other helplines between 1 and 24 times. Similar numbers of participants used the internet in both groups (seven in the BSC group vs. eight in the usual-care group), but the participants in the usual group appeared to use the internet more frequently (one or two times in 12 months in BSC group vs. 2–10 times in 12 months in usual-care group). Nine participants in each group used self-help booklet for advice to stop smoking. Except for one participant in the BSC group who reported using the booklet 40 times during the trial period, the remainder of participants used the self-help booklet fewer than ten times in the same period. These results could indicate an unmet need for smoking cessation support in this population.

The results from self-report in the follow-up questionnaire indicated that in both groups more participants were using smoking cessation products at 12 months than at baseline (36 in the BSC group vs. 21 in the usual-care group), but the proportion in the BSC group remained higher than in the usual-care group (78% vs. 41%). Comparing data extracted from GP records and participants’ self-report demonstrated that, while participants in the BSC group remained more likely to use pharmacotherapy than those in the usual-care group, participants in both groups tended to use pharmacotherapy obtained from other sources in addition to, or instead of, that obtained through GP prescription (Table 30). Considering that the participants in this sample rarely seek help through NHS Stop Smoking Services, it was reasonable to assume that products not covered by GP prescription were purchased over the counter. This also suggests that participants in the usual-care group tended to obtain their products over the counter. Overall, 42% of the pharmacotherapy cost was spent over the counter in the BSC group, whereas 66% was spent over the counter in the usual-care group. Furthermore, the results also indicated that participants in the BSC group were more likely to receive other NRT products (Figure 17). However, this observation was not evident when using self-report information. Although participants in the BSC group appeared to be more likely to receive multiple pharmacotherapies, the range of products used was similar in both groups (Figure 18).

TABLE 30 - Mean cost of pharmacotherapy during 12-month trial period

Trial arm	Pharmacotherapy cost per participant
	By prescription (SD), recorded by GP surgeries	By prescription and over the counter (SD), reported by participants
BSC	£62 (£132)	£106 (£138)
Usual care	£17 (£60)	£50 (£56)

FIGURE 17.

Number of participants using pharmacotherapy during trial period by GP records, by group and pharmacotherapy. UC, usual care.

FIGURE 18.

Number of participants using pharmacotherapy during trial period by self-report, by group and pharmacotherapy. UC, usual care.

Piloting the health-care and social care service use questionnaire

In order to explore the feasibility and scope of analysis full of a economic evaluation, a questionnaire was used to collect data on health-care resources and community services utilisation. While a comprehensive list of services available was essential to the analysis, the pilot identified several issues when using this questionnaire with this population.

Firstly, by compiling a complete service list, some of the community services were potentially overlapping. For example, the duties of a care co-ordinator, case manager and key worker could be carried out by a community psychiatric nurse (CPN) or a district nurse, since these positions are in some cases interchangeable between different staff members. Listing roles separately appeared to result in confusion and possible double counting. However, a lack of conclusive evidence means that we cannot determine if double counting occurred. Consequently, although we are aware of it, the results reported here were not adjusted for this possibility. It is something which would be addressed in a full trial.

Secondly, the groupings of services highlighted potential issues regarding costs. For instance, a health visitor could be a district nurse or other personnel, depending on the situation. The average cost per health visitor visit is higher than for a district nurse, so further distinctions need to be made. Although it might not cause problems on the participants’ side, the cost cannot be valued properly when the two are listed as one service.

In addition, some of the services are available from both public and private sectors. Without supervision, the nature of the self-report questionnaire might lead to a result that included the private sector, regardless of the intention. As the final two open questions regarding usage of other services indicated, although the questions were clearly listed under community service, some participants’ responses indicated that they might not have a clear idea of the nature of service providers. Some of the services mentioned were apparently not provided by the NHS or social care service, while others were provided by both private and public sector. Without further information, we were unable to determine what proportion of the services, if any, which were provided by public sector.

It should also be noted that, while we attempted to determine pharmacotherapy use from GP records in order to derive more accurate estimates, there were discrepancies between GP records and self-report which could not be fully explained by out-of-pocket purchase. This could result from recall error or participants failing to follow instructions. However, although prescription information collected from GP surgeries is believed to be more reliable, it is unlikely to be replicated in a full trial with a larger sample size. The workload could be too high for GP surgeries to respond.

Because of the nature of this population, any information collected through personal recall could be less reliable than under normal circumstances. Therefore, cost assessment regarding this population should be interpreted with caution.

Summary of findings

The incremental cost of providing the BSC intervention over and above usual care was estimated at £221 (SD £160) per participant. When the wider use of health-care and social care and prescriptions is included, the total cost in the BSC group was £12,674 (SD £16,596) per participant, compared with £6867 (SD £6026) per participant in the usual-care group. However, because of the small sample size, we recommend that these results should be treated with caution as the means are influenced by extreme values. Combining costs with the number of successful quitters at 12 months, the incremental cost was £58,197 per quitter. However, these results are from a pilot trial which is not powered to detect a significant difference from an economic perspective. It is also likely that in the longer term, beyond the 12-month follow-up, cost savings may accrue as a result of successful quits. Furthermore, improvements in health-related quality of life would be expected beyond the trial follow-up, which would be modelled in a full trial.

The pilot trial demonstrates the feasibility of conducting a full economic evaluation in a sufficiently powered trial of BSC over and above usual care. Several issues with regard to questionnaire design have been identified which would improve the accuracy and completeness in the collection of service use data.

Background/introduction

We explored specific issues of acceptability and engagement with the BSC interventions among patients with SMI and with professionals who delivered the intervention. An understanding of these issues is essential for improving the implementation of a BSC service and informing the design of the intervention in subsequent definitive trials. The aim of the substudy was to qualitatively explore, from both patient and therapist perspectives, perceptions of the need for smoking cessation services for this population and their experience of delivering or receiving the bespoke intervention. In particular, we aimed to explore how the bespoke intervention differed to any previous experiences of smoking cessation in usual care, and to identify barriers and facilitators of implementing or engaging with the intervention in practice.

Methods

In-depth semistructured interviews were conducted with a purposive sample including participants who completed the intervention and those who struggled to engage, and to compare those who sustained engagement with those who struggled or withdrew from treatment, and to identify barriers and facilitators to patient engagement. All participants in the intervention arm who had completed their treatment with the MHSCP or withdrawn from treatment were invited to take part. We did not preselect participants based on any other specific criteria, such as sex or smoking history. Participants who responded by post with an expression of interest or who verbally informed either their MHSCP or the research team that they were interested were contacted by telephone to discuss participation. We performed a comparison of the interviewed sample to the full trial sample on predetermined variables (including age, sex, ethnicity, number of previous quit attempts, smoking history and SMI diagnosis) to determine the representativeness of the interview sample compared with the patient sample as a whole.

We also conducted semistructured interviews with the MHSCPs to gain their perspectives on acceptability and delivery of smoking cessation services for SMI.

In-depth interview topic guides addressed the following issues:

• Characteristics of the recipients: what are the specific features of SMI that need to be anticipated and accommodated in delivering BSC?

• Mode and setting of delivery: is BSC best delivered in patients’ homes, GP surgeries or day-hospital settings? Is BSC best delivered face to face, in groups or over the telephone? What is an ideal contact time and number of sessions?

• Prior experience of smoking cessation, including support received from other primary care or mental health professionals.

• Acceptability of the intervention to patients, and satisfaction with the BSC, particularly in comparison with previous smoking cessation interventions received.

• Patients’ engagement with the intervention, with specific reference to barriers and facilitators to working with the MHSCPs.

• Implementation in routine care, including perceptions of who is best to deliver the BSC and any anticipated barriers to implementation.

Qualitative interviews were held at the end of treatment and ran throughout the duration of the data collection period. An experienced qualitative researcher facilitated all interviews and ethical approval was obtained by the relevant local NHS research ethics committee. Written consent was collected from all of the participants. Participants were asked to consent to the discussions or interviews being recorded and were informed that all identifiable data would be removed once transcribed. Participants were informed that they could remove themselves from the group or stop the interview at any time and did not have to answer any questions they were uncomfortable with.

After completion of the interview, as a token of thanks for their time, the participant was offered a £10 gift voucher as a good-will gesture.

Changes to the original protocol

In the original protocol we intended to interview practice staff who had been involved in the delivery of the intervention (GPs, practice nurses). However, during the study it became apparent that such staff had minimal involvement in the intervention itself and, therefore, we did not interview GPs directly. We did, however, modify the topic guides for both patients and MHSCPs to ask specifically about their interactions with GPs to ensure any relevant issues were captured (and to reconsider the need for interviewing GPs if it became apparent that their involvement was greater than expected, although the interviews confirmed our perceptions that GPs had minimal involvement in delivering or referring to the intervention). The following results are, therefore, from the patient and MHSCP interviews.

Participant characteristics

Interviews took place between August 2012 and January 2013. Thirteen patients were recruited from across the three recruitment sites (five from Manchester, six from York and two from Hull) and three MHSCPs, one from each site, were interviewed.

Analysis

Each in-depth interview was digitally recorded and transcribed verbatim. Transcripts were checked and anonymised to remove identifying details.

Transcripts were read independently by two researchers and analysed using the constant comparison (CC) method. The CC method aims to inductively develop themes through categorising and coding data and exploring connections between them, repeating the cycle across the data set until theoretical saturation is achieved. Emergent themes were discussed and verified with a third researcher. Analysis was completed prior to the quantitative analysis being complete, and was therefore blind to study outcome.

The MHSCP transcripts were initially analysed independently from the patient transcripts, but the analysis was combined when preliminary readings suggested consensus in core themes across both the patient and professional data sets and also indicated that novel insights could be synthesised across the two samples to provide a holistic picture of the intervention. Similarly, we did not analyse the data of engaged and disengaged patients separately, partly because only two of the participants were formally considered to have ‘disengaged’ (having been discharged because of a lack of contact with the MHSCP), but also because analysis suggested this separation did not reflect differences in experienced barriers and facilitators as we has assumed it would. Both the disengaged patients reported positive experiences with the intervention itself, suggesting other circumstances may have contributed to their disengagement, and even patients considered to have engaged with treatment reported difficulties maintaining motivation and planning future sessions. This suggests that engagement was less reflective of the acceptability of the intervention and more indicative of the chaotic nature of this population, which was a recurrent theme in the data (theme 3 below).

Main findings

We identified four primary themes. Themes 1 and 2 reflected the lack of support for smoking cessation in current services and, consequently, the perceived benefits of the BSC intervention which was more tailored to this population. Themes 3 and 4 reflect challenges and barriers reported by patients and professionals, including difficulties sustaining engagement and difficulties liaising with primary care.

Summary of themes

Overall, the findings of the qualitative substudy support the need for a sensitised and BSC intervention for this population. Providing this through mental health trained workers was perceived to be most appropriate by both patients and professionals. Challenges to be addressed include difficulties in helping patients to manage their cessation plans, and better communication or integration with primary care to organise prescribing. Implementation in routine care settings, particularly considering who would take on the MHSCP role and cost implications of this, should also be explored.

Limitations

Only three MHSCPs could be interviewed given that this was a pilot study. Future work should explore whether or not larger cohorts of MHSCPs report similar experiences. Although the patient sample is fairly small, the participants interviewed included patients with both bipolar disorder and schizophrenia and were largely representative of the overall trial population. The consistency in emergent themes across the patients also supports the representativeness of the results. However, women and black and minority ethnic participants were under-represented and further research is necessary to explore the acceptability of the BSC intervention to this population.

Main findings

The main finding of the SCIMITAR trial is that smoking cessation can be achieved among people with SMI and that the use of a BSC intervention increased the chances of sustained quitting, as estimated by a biochemically verified outcome measure (exhaled CO). 72 The observed odds of successful quitting at 12 months were almost three times higher among those who received BSC (OR 2.9, 95% CI 0.8 to 10.5), although this value was calculated using a small sample and, therefore, requires cautious interpretation.

A range of secondary outcomes were also measured and there was a general direction of effect in favour of BSC in relation to the (1) number of cigarettes smoked, (2) number of quit attempts, (3) reported nicotine dependence and (4) reported MTQ. Taken together, the positive overall primary outcome and consistency of direction of effect among primary and secondary outcomes (reduced number of cigarettes smoked, increased number of quit attempts, increased MTQ) add weight to the hypothesis that BSC is effective for this group. The consistency of findings from the pilot trial, alongside systematic review evidence44 represents an accumulation of evidence. Ultimately the clinical effectiveness of a BSC intervention can really be tested only within a fully powered RCT. The pilot trial also found some evidence of deterioration in mental health in the intervention group compared with the usual-care group. This finding is not consistent with other evidence, where smoking cessation tends to improve mental health. 73 However, this does indicate that mood should be monitored in clinical practice and the safety of smoking cessation should be tested in a fully powered trial. The various findings and experience from conducting a pilot trial will now be considered in turn in order to inform the design of a fully powered definitive RCT.

Conclusions

The SCIMITAR pilot trial has shown that it is possible to recruit to a trial of a BSC intervention for people with SMI. Follow-up in 70% of participants has been achieved using a biologically verified measure smoking status at 12 months. The preliminary estimates of clinical effectiveness are supportive of BSC across a range of primary and secondary outcomes. The clinical effectiveness and cost-effectiveness of bespoke smoking can now be established in a fully powered trial. A recruitment strategy for a fully powered trial should enlist participants from primary and secondary care and the SCIMITAR pilot trial has delineated the relative strengths and practical limitations of approaches in both of these settings. Both approaches should be used in a definitive trial.

Contribution of authors

Simon Gilbody and Mei-See Man wrote the original protocol.

Simon Gilbody, Jinshou Li, Susan Michie and Tim Bradshaw were co-applicants on the Health Technology Assessment application and refined the protocol.

Simon Gilbody was the chief investigator and oversaw the study.

Mei-See Man, Natasha Mitchell and Emily Peckham were trial mangers.

Taeko Becque designed and conducted the clinical analysis.

Jinshou Li and Steve Parrott designed and undertook the economic analysis.

The writing team consisted of Taeko Becque, Simon Gilbody, Sarah Knowles, Mei-See Man, Emily Peckham, Claire Planner and Charles Shepherd who drafted the report.

Collaborations

The SCIMITAR collaborators (current and past) are:

Katie Atherton, Taeko Becque, Tim J Bradshaw, Helen Cox, Ben Cross, Jane Dallender, Emma Davies, Rhian Gabe, Linda Gask, Simon Gilbody, Edward Greenward, Kerin Hannon, Laura Hermann, Catherine Hewitt, Hayley Jackson, Sarah Knowles, Helen Lester, Jinshou Li, Andy McEwen, Mei-See Man, Sarah Mercer, Susan Michie, Natasha Mitchell, Ann Mortimer, Emily Peckham, Claire Planner, David Richards, Kath Richardson, Charles Shepherd, Maggie Stronach, David Torgerson, Muhammad Usman, Ian Watt and Robert West.

Trial Steering Committee members

Professor Ann McNeil (independent chairperson), Professor of Tobacco Addiction, King’s College London, London, UK.

Dr David Shiers, retired GP, North Staffordshire.

Dr Tom Hughes, Consultant psychiatrist, Leeds and York Partnership NHS Foundation Trust, Leeds, UK.

Data Monitoring and Ethics Committee members

Dr David Osborn (independent chairperson), Reader in Community Psychiatric Epidemiology, Mental Health Science Unit, University College London, London, UK.

Dr Elena Ratschen, Lecturer in Epidemiology/Tobacco Control, Faculty of Medicine and Health Sciences, University of Nottingham, Nottingham, UK.

Dr Richard Emsley, Lecturer in Biostatistics, Centre for Biostatistics, University of Manchester, Manchester, UK.

Patient and public involvement in research

The SCIMITAR trial benefited from involvement of users of mental health services and carers of people with SMI throughout the research period. Our Trial Steering Committee included representation from a carer. Our protocol and study materials were scrutinised and supported by users and carers in the north-west of England and by the user and carer groups of our local MHRN.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health.

Publications

Bradshaw T, Davies E, Stronach M, Hermann L. Support for people with serious mental illness to cut down or stop smoking. Mental Health Pract 2014;17:14–20. URL: http://dx.doi.org/10.7748/mhp2014.03.17.6.14.e890

Gilbody S, Peckham E, Man M-S, Mitchell N, Li J, Becque T, et al. Bespoke smoking cessation for people with severe mental ill health (SCIMITAR): a pilot randomised controlled trial. Lancet Psychiatry 2015; in press. http://dx.doi.org/10.1016/S2215-0366(15)00091-7

Suicide protocol

(PDF download)

Adverse event protocol

(PDF download)