Tailored tobacco dependence support for mental health patients: a model for inpatient and community services
Correspondence to: Elena Ratschen, Division of Epidemiology and Public Health, UK Centre for Tobacco Control Studies, University of Nottingham, Nottingham NG5 1PB, UK. E-mail: firstname.lastname@example.org
Although smoking prevalence among people with severe mental illness is high, it remains largely unaddressed. This pragmatic pilot project aimed to develop and implement a tailored tobacco dependence service in mental health settings and to assess its impact, as well as barriers and facilitators to implementation.
An integrative service model, spanning acute, rehabilitation and community services, including the design of tailored instruments and referral pathways, delivered by two mental health professionals.
Setting and participants
Four adult acute and two rehabilitation wards (129 beds), and the community recovery team (2038 cases) of the United Kingdom's largest Mental Health Trust.
Audit of smoking information in patient notes; service uptake; quit attempts; smoking cessation and reduction; qualitative data on implementation barriers/facilitators.
A total of 110 patients attended at least one support session: 53 inpatients (23% of inpatient smokers) and 57 community (of unknown number of community smokers, as recording of smoking status is not mandatory). Thirty-four of these (31%) made a quit attempt; 17 (15%) stopped smoking and 29 (26%) reduced cigarette consumption by up to 50% at the final contact. Barriers to service implementation related to: (i) trust policy, systems and procedures, (ii) staff knowledge and attitudes and (iii) illness-related factors.
Despite the strong anti-smoking climate in the United Kingdom, including a law requiring smoke-free policies in mental health settings, establishing a smoking cessation treatment service for people with mental illness proved difficult, due to complex systemic barriers. However, there is clearly a demand, by patients, for such a service.
People with severe mental illness (SMI) are up to three times more likely to be smokers than the general population [1, 2], are often heavy smokers and are more severely dependent [3, 4]. Consequently, people with mental health disorders have worse physical health than the general population, and the English government's Tobacco Control and Mental Health Strategies have made it a priority to tackle smoking in people with mental health problems [5, 6].
The underlying reasons for the strong relationship between smoking and mental illness are complex, and include neurobiological, psychosocial and genetic factors [7, 8]. Smoking often constitutes a means of social interaction, reducing social inhibition and isolation encountered frequently in this population, and relieving boredom in treatment settings [7, 9]. Smoking also induces liver enzymes responsible for clearance of some psychiatric medication, thus increasing drug clearance and necessitating higher medication doses .
However, smoking is rarely addressed in mental health settings, and although smoking has been banned in indoor areas since the introduction of smoke-free policies in mental health settings in 2007, it remains deeply embedded in the culture [11, 12], with clinicians reluctant at times to address smoking as an integral part of treatment [11-13]. While a societal change towards a decreased acceptance of smoking has taken place in the United Kingdom over recent years, resulting in a relatively strong anti-smoking climate , it is still largely condoned across psychiatric settings, and many mental health professionals perceive it as an important coping mechanism for patients [12, 15].
Contrary to common perception, patients with SMI are frequently willing  and able to quit smoking provided they receive appropriate support . However, smoking cessation support delivered through National Health Service (NHS) Stop Smoking Services (SSS) is standardized through guidance and training and tailoring to individual circumstances and needs is not actively encouraged or widely employed. In addition, support is provided by advisers who typically do not have the necessary skills and experience to deal with smokers with mental health problems. Pharmacological treatment with both nicotine replacement therapy (NRT) and bupropion, given separately or in combination, has proved effective and well tolerated in psychiatric populations [17, 18] (the most recent pharmacological treatment, varenicline, is currently being trialled for safety in the psychiatric population). Given the high levels of nicotine dependence, smoking reduction may be a good way to achieve eventual abstinence . A recent Cochrane Review  found, among general population smokers, that quit rates were similar whether quitters used an abrupt quit or a reduction to quit intervention when behavioural support was provided, suggesting that smokers can be given the opportunity to quit either way. A further review indicated that NRT can be used to support reduction attempts in those smokers not willing or able to quit smoking abruptly .
There are no current recommendations on the best way to provide a comprehensive programme for treating tobacco dependence in mental health settings. The aim of this pragmatic pilot project was to develop and implement a tailored tobacco dependence service in mental health settings and to assess its impact, as well as barriers and facilitators to implementation.
The pilot project was carried out within adult mental health treatment services in the United Kingdom's largest Mental Health Trust from October 2010 to June 2011. Of the seven acute mental health inpatient wards potentially suitable for piloting the programme, two were excluded because they had served as study pilot wards in previous research (2008/09) [22, 23]. Two of the remaining five acute inpatient wards and two of four rehabilitation settings, where patients receive care with the aim of reintegration into the community, were chosen at random, resulting in an initial sample of four wards (85 beds). Seven months into the study, two more acute wards (44 beds) were incorporated because the ward managers expressed an interest in participating (total of beds = 129). The recovery team was chosen to participate as a community-based service because it gave care to the largest number of patients of all community teams (2038 cases). All patients were eligible for inclusion if they smoked and wanted to address their smoking, and—in the case of inpatients—were stable enough as assessed by clinical staff.
Instruments and procedures
At the beginning of the project, an audit of current procedures related to smoking was carried out in the four wards and recovery team. Relevant policies, such as the Trust's smoke-free policy and any information relating to the treatment of smoking within patients' care pathways, were reviewed and any documentation of smoking status and prescription of NRT was recorded from the electronic patient information system used by the Trust for all current inpatients and community patients, and from paper notes for all current inpatients and for a random sample of 50 community patients.
For the pilot project, two mental health professionals trained in smoking cessation worked full time within the sample settings to support patient and staff smokers to follow structured quit and assisted reduction programmes. The quit programme was evidence-based and followed the standard NHS SSS treatment, but was designed to provide flexible support tailored to the needs of individual patients and delivered by health professionals who were experienced and knowledgeable in treating people with mental health problems. The programme provided NRT that was available on the Trust formulary, thus avoiding the complexities involved in prescribing either the anti-depressant bupropion, or varenicline, which currently carry a caution for use in this population, and may not have been well supported by clinicians. It involved setting individual goals and providing patients with techniques to achieve them. The reduction programme involved working towards reducing cigarette consumption by 50% , offering NRT with the ultimate aim or prospect of quitting. The approach was aimed at clients who were not ready for, or did not have the confidence to quit straight away, but expressed an interest in addressing their smoking. The service included motivational interviewing, cognitive behaviour therapy (CBT) techniques, combination NRT and tailored materials for patients. This was delivered flexibly in both 1 : 1 and group formats, allowing weekly monitoring of goals and the duration and intensity of support to be adapted to individual needs and motivation. For inpatients, the support was delivered by an adviser on site, whereas in the community the adviser offered clinics in easily accessible places (e.g. community centres) or, in liaison with key workers, visited patients at home for support. Patients were recruited through direct contact with the advisers, with the help of inpatient and community psychiatric nurses (CPNs) and, only in the community, through a mail-drop to all patients advertising the service. Staff smokers were offered support opportunistically.
The team developed assessment sheets and care pathway documents for the advisers and staff to use. They also set up referral and communication pathways for smokers, including an online referral system to the local NHS SSS. The advisers wrote letters to general practitioners and the psychiatric teams to update them on patients' progress in tackling their smoking. All members of staff in the participating settings were offered brief intervention training and, on completion, those interested were trained as smoking cessation advisers. The aim was to ensure that staff stayed up to date with the project, understood guidelines on NRT and smoking and monitored changes in psychotrophic drugs as needed.
For the pilot project, the following measures were collected.
Demographic detail, smoking status and, wherever possible, diagnosis as per case notes, were recorded for all patients engaging with the service.
The number of patients who attended support sessions with the advisers, the total number of facilitated individual and group support sessions and referrals to the local NHS SSS were recorded.
Smoking-related outcomes referred to measures of smoking abstinence and reduction in both patients and staff. Point-prevalence abstinence was defined by self-report and expired air carbon monoxide (CO) reading at last patient contact, with CO readings of 10 parts per million or lower deemed to indicate abstinence . A successful assisted reduction outcome was defined by maintained reduction at 50% for at least 2 weeks prior to end of support. Information on reasons for failed quit/reduction attempts were also recorded. Analyses were conducted in SPSS, using descriptive statistics.
Barriers and facilitators to implementation
Both advisers collected information on barriers and facilitators encountered in each setting on a weekly basis, using structured recording sheets. These were discussed with the project team in bi-weekly meetings, to agree points of action to be taken to address these. Two members of the research team grouped the main issues identified into themes to facilitate structured reporting and conceptual understanding.
There was no Trust policy on smoking dependence treatment; however, the Trust smoke-free policy stated that its aim was: ‘to encourage an environment conducive to giving up smoking, which provides full support to staff and patients who want to give up smoking’. The audit revealed that smoking status was a mandatory recording item only for inpatients, and there were no standard procedures related to recording the provision of smoking-related advice and support in any of the settings sampled. Of the inpatients audited, 73% were current smokers, but only 24% had received recorded advice on the risks of smoking (Table 1), and only one inpatient had been referred to the NHS SSS. Of the 2028 community patients, only 450 (22%) had an electronic record of their smoking status; of these, 274 (61%) were identified as smokers.
Table 1. Audit—smoking-related information available from inpatient case-notes.
|Acute ward 1||18||11 (61%)||0||0||0|
|Acute ward 2||22||15 (68%)||2 (13%)||0||0|
|Rehabilitation ward 1||18||13 (72%)||4 (31%)||1 (8%)||1 (8%)|
|Rehabilitation ward 2||27||23 (85%)||9 (39%)||0||1 (4%)|
|Total||85||62 (73%)||15 (24%)||1 (2%)||2 (3%)|
Of the inpatients (acute and rehabilitation wards) available for treatment throughout the pilot period, 243 (63%) were smokers, all of whom were approached, given advice on smoking and offered tobacco dependence support. Overall, 57 patients (23%) engaged in at least one 1 : 1 support session (Table 2). Group support was not utilized, as all inpatients preferred an individual format. Outcomes are displayed in Table 3.
Table 2. Client characteristics of those engaging with support.
|From acute wards||40 (70%)||Not applicable|
|Male||36 (63%)||31 (58%)|
|Mean age (years)||44 (range 21–65)||49 (range 30–64)|
|Number of patients who reported previous quit attempts||36 (63%)||41 (77%)|
|Mean daily cigarette consumption||15 (range 0–40)||32 (range 0–60)|
|Diagnosis||Depression||18 (32%)||12 (23%)|
|Schizophrenia||16 (28%)||15 (28%)|
|Bipolar disorder||8 (14%)||7 (13%)|
|Other||15 (26%)||19 (37%)|
Table 3. Smoking outcomes for those engaging with support.
|Total no. individual appointments attended||186||125|
|Total no. group sessions held||Not applicable||12|
|Average (mean) no. appointments/patient||3||3 (median 2)|
|No. of quit attempts made||11 (19%)||23 (43%)|
|No. of patients abstinent at end of support period||4 (7%)||13 (17%)|
|No. of 50% reductions||4 (7%)||1 (2%)|
|No. less than 50% reductions||14 (25%)||10 (19%)|
The option of utilizing NRT to aid quit and reduction attempts was well received, with 27 patients (47% of patients who attended at least one appointment) opting to have NRT prescribed. All patients who attempted to quit utilized NRT; however, although combination NRT was promoted, all four inpatients who quit successfully utilized only one NRT product (two used 21-mg patches; two used 10-mg inhalators).
The length of abstinence achieved during failed quit attempts ranged from 1 to 5 weeks (mean = 3 weeks), with reasons given for relapse including stress during the Christmas/New Year period, stress related to physical illness and giving in to cravings.
Discharges into the community (n = 20, 35%) and patients changing their mind about addressing smoking (n = 20, 35%) were the most common reasons for ending support. Six patients who were discharged and wished to address smoking in the community were referred to the community adviser.
Following a letter to all patients and briefing of CPNs the community adviser received 75 referrals, of which 32 (43%) were self-referrals. Fifty-three (70%) of these patients were contacted successfully (after a median of six attempts) and had at least one appointment with the adviser. Reasons for the 22 referrals not translating into appointments included patients changing their mind about addressing smoking (n = 7, 32%) and the adviser being unable to establish contact (n = 13, 59%). For all patients, health (47%) and financial (13%) concerns were the most common motivations for accessing the service this time. Table 1 shows client characteristics.
The main format of support was individual 1 : 1 sessions, although some opted for group support and two groups of six sessions were run during the course of the project with an average of two patients in attendance. Patients preferred support from the mental health specialist adviser, with only seven referrals chosen to NHS SSS over 9 months, none of which translated into actual attendance at the service.
A total of 24 patients (45%) opted to use NRT, obtained from their general practitioner (n = 14) via the NHS SSS (n = 5) or pharmacies (n = 5). The most frequently used products were 4-mg nicotine gum (n = 13) and 21-mg patches (n = 11). There was no significant difference in the number of patients using NRT between the successful and unsuccessful quit attempt groups (P > 0.05, Fisher's exact test).
The most common reasons for end of engagement with the project were patients no longer wishing to address smoking (n = 20 patients, 38%) and loss to follow-up (n = 15, 28%).
With the advisers' support, eight staff members made a quit attempt and four succeeded. One staff member reduced by 50%, four by less than 50% and the remainder received information on how to access NHS SSS and prepare for a quit attempt in the future.
Barriers and facilitators
A number of barriers to implementing tobacco dependence support were encountered during the project. These were subsumed under three themes: (i) Trust policy, systems and procedures, (ii) knowledge, skills and attitudes and (iii) illness-related factors, as discussed below.
Trust policy, systems and procedures
Procedures relating to the regular facilitation of smoking varied across the settings. In places, they allowed staff to smoke with patients, and patients to smoke on Trust premises, whereas the smoke-free policy stated that Trust premises, indoors and outdoors, should be smoke-free. There was a lack of resources to achieve the goals outlined in the smoke-free policy, such as tobacco dependence treatment guidelines, bespoke recording instruments and NRT stock available to staff to implement smoking cessation support. No targets for reducing smoking or treating smokers beyond the recording of smoking status in inpatient settings could be identified.
Knowledge, skills and attitudes
Staff had received basic smoking-related training through an online e-learning module, which encompassed smoking facts in relation to physical and mental health. During the implementation of tobacco dependence support, it became apparent that there were shortfalls in staff knowledge and negative attitudes towards quitting. These included concerns about the ‘harmful effects’ and expense to the Trust of NRT, patients' social isolation following cessation, smoking being an important patient coping strategy and concerns that ‘it was not the right time’ for patients to quit even if they wanted to.
Both the inpatient and community settings provided services to patients experiencing severe and enduring mental illness, and therefore significant barriers were related to attentional, cognitive and motivational factors, sometimes resulting in difficulties to engage and retain patients.
A number of steps were taken to overcome barriers related to the themes detailed above, including comprehensive staff training, dissemination of the audit and interim project results, development of recording instruments and collaborative pathways, close liaison with management and consultants and a flexible, responsive approach to patients' needs, including the integration of a peer-support element through the development of ‘quit stories’ by successful quitters. These actions were deemed by the project advisers who were working full time within the sample settings to raise the profile and acceptance of tobacco dependence support during the study.
Smoking prevalence in the sample settings was found to be nearly three times higher than in the general population (63% of inpatients; 61% of community patients for whom smoking status was recorded), and patients tended to be heavy smokers, with community smokers consuming an average of 32 cigarettes daily. Despite these high levels of smoking, the audit revealed that smoking status was a mandatory recording item only for inpatients and there were no standard procedures related to recording the provision of smoking-related advice and support in any of the settings sampled. A total of 110 patients, 57 from inpatient and 53 from community settings, engaged with the service and attended at least one appointment with the specialist advisers, 34 patients (31% of 110) made a quit attempt, 17 (15%) had stopped smoking at the end of their individual support period, and a further 29 patients (26%) reduced their cigarette consumption by 50% or less. Nine staff altered their smoking behaviour. These findings indicate modest success in terms of absolute numbers of successful quit attempts, but reflect an interest in and demand for support. They also indicate, in line with translational research and implementation science , that there is a substantial discrepancy between outcomes of efficacy studies conducted in controlled settings (such as those included in a recent systematic review ) and those following the translation of the evidence base into everyday service provision. To develop and deliver tailored, ‘workable’ support models in mental health settings successfully, complex systemic barriers remain to be addressed, even in a country with a strong anti-smoking climate such as the United Kingdom. These include addressing staff knowledge, attitudes and behaviour related to smoking (including staff smoking), the development of appropriate documentation systems and pathways to record smoking-related information and provide appropriate support.
There were several methodological and organizational limitations to this project. This study was designed as a pragmatic project, as time and resource constraints did not permit a randomized controlled design. It is therefore unclear what would have happened in these settings in the absence of these interventions. However, based on the audit conducted at baseline, which revealed a lack of support structures, as well as on previous research in similar settings of the Trust that were excluded from this project , it seems reasonable to assume that due to a lack of information, encouragement and support very few, if any, patients would have quit or reduced their smoking in the absence of the service offered in the context of the project. Logistic limitations were related mainly to the development of prescription pathways for NRT in the community, but we believed that restricting the pilot to the use of NRT (rather than also supplying varenicline or bupropion) helped to smooth the way to implementation of the programme. Furthermore, despite repeated efforts, psychiatric consultants gave limited support to the pilot. Different ways to engage this group, for example through mandatory training, require research.
While the rate of service uptake appears modest, it should be emphasized that in the inpatient setting, where the adviser was able to recruit patients through direct contact, almost a quarter of all smokers (23%) engaged with the service, although this setting had been anticipated to yield low numbers of engagement due to the severity of patients' mental health conditions. Patient recruitment in the community was made difficult, given that smoking status was not recorded for the majority of community patients. Recording of smoking status in the community needs to be made mandatory. We believe that if smoking status had been recorded, this would have enabled advisers and community staff to approach and follow-up patient smokers directly, resulting in improved service uptake and quit rates. However, although low in absolute terms, the rate of quit attempts and successful quits and reductions are encouraging, given the particular difficulties (such high nicotine dependence, complex life circumstances) faced by smokers in this population. Previous research conducted internationally and within UK mental health settings had found a persistent smoking culture, with no clear strategies and appropriate resources to address smoking, despite the implementation of smoke-free policies in mental health settings [12, 22, 26, 27]. This was also found to be the case in the pilot sample settings, with little guidance and training for staff and negative staff attitudes prevailing. Specific situations encountered during this project, such as staff advising patients against quitting or expressing negative attitudes towards NRT use, staff smoking with patients and smoking being the predominant activity on wards, support evidence put forward by papers that call for change [12, 22, 23, 26, 27]. The barriers and future facilitators of service provision identified are similar to those described for US and Australian settings [11, 27, 28], and indicate the need for a comprehensive approach towards changing the smoking culture.
Feedback suggested that tailoring the duration and intensity of counselling to patients' needs was well received by patients and staff. It became evident that in many cases, flexibility of support was essential. For example, in one particular case, a patient attended 17 sessions before feeling confident in their ability to maintain abstinence without support. Arguably, specialist mental health settings are well placed to provide appropriately tailored support through mental health specialist advisers, who could rotate between different settings with protected time, or on a full-time basis to provide a service to patients and staff, as well as training for the latter. This seems to be particularly important in light of the finding that some patients in this project did not wish to engage with the local SSS.
Review evidence  suggests that ‘Cutting Down to Quit’ support could help those with SMI who are not ready to quit abruptly. In this project, overall 29 patients (26%) reduced as a first step towards quitting with structured assistance. Although we were not able to demonstrate that assisted reduction resulted in quit attempts, because follow-up periods were too short, patients reported that reducing cigarette consumption had had a positive impact on their finances and improved confidence in their ability to quit smoking in the future, indicating that trial evidence to assess the effectiveness of ‘Cutting Down To Quit’ on long-term smoking outcomes in this population is needed. Our results also support previous findings of high relapse rates in this population, therefore research into effective relapse prevention strategies would be of benefit.
Results from this pragmatic pilot study highlight the difficulty of implementing evidence-based tobacco dependence treatment in service settings with powerful smoking cultures, even in a country such as the United Kingdom with advanced tobacco control policies and smoke-free policies in mental health settings. In the context of the numerous systemic barriers encountered, and in view of the particular difficulties encountered by many people with SMI to address their smoking, service uptake and quitting/reduction outcomes were, albeit low in absolute numbers, encouraging in relative terms.
Barriers to service implementation identified suggest that the employment of dedicated staff to provide the service, for example by rotating through a variety of mental health inpatient and community settings and providing staff training, may be more successful than relying on the delivery of a comprehensive service by regular staff without protected time, or the NHS SSS. Strategies to ensure that smoking is being recognized as a matter of importance by clinicians and managers are required urgently. For example, the mandatory recording of smoking status for all patients in contact with mental health services needs to be put in place urgently, as this is a prerequisite for the appropriate identification of smokers and subsequent offers of support. Also required are the implementation of comprehensive smoking-specific care pathways and the development of financial incentives in collaboration with commissioners. These could include smoking-related targets and performance indicators, set by commissioners for Mental Health Trusts, that consider clinical performance and activity beyond mere recordings of smoking status, such as documentation of smoking history, dependence and quitting behaviour, and provision of brief advice and treatment support according to integrated smoking care pathways. These should be supported by mandatory training related to smoking and its specific links with mental illness, for all staff, including junior doctors and consultants. Support needs to be tailored and more flexible in terms of duration and intensity and standard NHS SSS monitoring and success criteria, such as measuring 4-week quit rates, are not appropriate and need broadening for this population. Further research with long-term follow-up to explore Cut Down To Quit and relapse prevention strategies are needed.
Declarations of interest
This research was carried out as part of a programme of inequalities projects implemented by the UK Centre for Tobacco Control Studies and funded by the Department of Health. The authors have no other conflicts of interest to declare.