Mental health units in England had to become smoke-free by law from July 2008. Concerns regarding the implementation and enforcement of smoke-free policies in these settings have been raised.
To study difficulties and challenges associated with smoke-free policy implementation in English National Health Service (NHS) mental health settings.
Questionnaire survey of all 72 English NHS trusts providing mental health in-patient services and facilities, supplemented by semi-structured telephone interviews at a systematic sample of 7 trusts and site visits at a convenience sample of 5 trusts.
Questionnaires were returned by 79% of the trusts, all of whom had implemented smoke-free policies. Most respondents (91%) believed that mental health settings faced particular challenges, arising from the high smoking prevalence among patients (81%), related safety risks (70%), adverse effects on the clinician–patient relationship (36%), and potential interactions with antipsychotic medication (34%). Interviews indicated that sustained policy enforcement was perceived as difficult, but that despite challenges and concerns, the impact of the policy was regarded as beneficial, with some evidence of positive behavioural changes occurring in people.
Many mental health trusts across England have implemented comprehensive smoke-free policies but the majority state that they are facing specific difficulties. Challenges and concerns need to be explored in depth and addressed to ensure that smoke-free policies implemented under the terms of the Health Act in July 2008 are not undermined.
Whereas National Health Service (NHS) acute trusts in England had to become smoke-free under the terms of the Health Act in July 2007, residential mental health units were given an additional year to comply, during which time they were allowed to designate either individual bedrooms or rooms to be used only for smoking for use by persons over 18 years of age.1 Complexities involved in the area of smoking and severe mental illness, for example high levels of smoking prevalence and nicotine dependence,2,3 as well as potential interactions between smoking and certain antipsychotic medication,4 have marked smoke-free policy implementation in mental health settings as an issue of considerable debate.5–7 Challenges, especially over sustained enforcement and staff concerns related to safety issues have been anticipated.8–11 In an earlier, unpublished survey by the Health Development Agency (details available from the author on request) of NHS acute and mental health hospitals conducted in 2003, the majority of respondents doubted that comprehensive smoke-free policies, including both buildings and grounds, could be implemented successfully in the NHS, with only 10% operating such policies at the time. In 2005, the Health Development Agency published Guidance for Smokefree Hospital Trusts12 to support trusts in implementing comprehensive policies. As part of a larger study of all acute and mental health trusts in England carried out between February and May 2007, we recently reported that all responding NHS mental health trusts had implemented smoke-free policies, 63% of which included provisions for buildings and grounds. However, exemptions were frequent (78%), over a third (37%) reported that infringements occurred at least daily and over 40% still maintained smoking rooms at the time of questionnaire completion.13 This paper provides further information from that study on specific concerns related to smoke-free policy implementation in mental health settings. Ethics approval for the study was granted by the Nottingham Research Ethics Committee (REC reference number: 07/Q2404/1).
The methods for this study have been described in detail elsewhere;13 the following paragraphs mainly focus on mental health specific aspects of the research.
Study institutions and participants
A list of all English NHS mental health trusts was purchased from the data provider Binley's, health and care information specialist, in January 2007. A total of 72 trusts were identified, including 17 primary care trusts that were listed as main in-patient mental health service providers for their region. Human resource directors who were most likely to be involved in the development of smoke-free policies within their trusts (details available from the author on request) were chosen as potential study participants.
Mental health specific measures of the study included the proportion of mental health trusts perceiving particular difficulties with smoke-free policy implementation and the nature of these difficulties; details on policy development, communication and support; and reported challenges and impacts of smoke-free policy.
The questionnaire was designed on the basis of previous research (details available from the author on request) and current guidelines12 to collect structured information on the development and implementation of trusts' smoke-free policies, and included a section on issues of specific relevance for mental health settings. It was sent to potential participants of all 72 trusts by post and additionally made accessible for online completion. Reminder letters and a request under the environmental information regulations were issued.13
Telephone interviews were carried out to supplement the information provided in the questionnaires with qualitative information and to identify issues related to perceived challenges and impacts of policy implementation. A semi-structured interview guide was developed for the interviews, allowing for the allocation of responses in predefined categories determined on the basis of previous research (details available from the author on request), and for the identification of further emerging themes that could inform further research. Interviews lasted around 30 min. Of the 86 questionnaire respondents (46%) who had indicated their availability for an interview, 30% systematic samples, stratified according to trust type, were drawn, resulting in a sample of 17 respondents from acute trusts and of 7 respondents from mental health trusts.
Site visits were carried out in a convenience sample of five trust sites to investigate visible indicators of smoke-free policy implementation and to triangulate data where possible.
Questionnaire responses were coded and entered into SPSS 14.0 for Windows database to generate the outcome measures. Telephone interviews were tape-recorded and responses allocated to predefined and emerging response categories. Selected verbatim quotes to illustrate respondents' perspectives are identified by participant number in the results section. Information from site visits was recorded in a checklist comprising information on signage and information related to smoke-free policy, and on smoking-related issues observed on the premises.
After excluding four questionnaires which revealed that, contrary to the database information, the responding primary care trusts did not provide mental health in-patient facilities, 79% of trusts approached participated in the study, consisting of 48 mental health trusts (87% response rate) and 6 (46%) of the primary healthcare trusts providing mental health in-patient facilities. Proportions of responses to first and follow-up letters, and to the environmental information regulations request have been detailed elsewhere.13 Almost half (46%) had indicated willingness to participate in a telephone interview and all seven respondents sampled were interviewed after obtaining informed consent. Site visits to publicly accessible areas of five trusts spread over three English regions, four of which had completed the questionnaire, were carried out.
Challenges and concerns
Ninety-one per cent of questionnaire respondents agreed that psychiatric settings encountered specific problems with regard to smoke-free policy implementation. In particular, respondents believed that the high prevalence of smoking among service users (81%) and concomitant safety issues (70%) were of concern. Adverse effects of the smoke-free policy on the clinician–patient relationship, problems related to the dosage of antipsychotic medication in the context of changed smoking behaviour and aggravation of mental health problems were believed to pose difficulties by 36%, 34% and 17% of respondents respectively.
Finding a way for staff to deal with aggression and abuse in the context of challenging smoking patients and visitors on trust premises emerged as a major concern during four interviews. The three remaining respondents, however, highlighted that anxieties related to increased aggression and incidences had proven unfounded.
`They [fears regarding violence and aggression] are unfounded. Having
spoken to other trusts that tried this [smoke-free policy implementation]
before – they are coming to the same conclusion. Everywhere you go, it's “
oh, it's gonna be awful, there will be violence and problems...” –
we have found that there is no increase in violence and aggression.'
Three respondents pointed out that the smoke-free policy was feared to increase fire risks, for example related to `underground smoking' in bedrooms. However, none reported the actual occurrence of such cases. One interview respondent commented on issues related to smoking regulations for forensic patients who had to be taken outside in secure areas as a result of the new policy, an arrangement that incurred practical difficulties and required considerable resources. Another reported that secluded patients were not able to smoke cigarettes anymore but had to take nicotine replacement therapy for the duration of their seclusion instead, and a third stated that the trust was about to adopt the same strategy for the forensic department.
The extensiveness and heterogeneity of trust sites emerged as a further issue of importance. Questionnaire responses show that all trusts comprised multiple sites. One interview respondent reported that the trust had decided not to include outdoor areas in the policy, since extensive grounds impeded effective policy enforcement. Two further interview respondents, one of them working at a trust with more than 100 sites, explained that coherence and consistency of policies in covering a vast number of different sites in different urban and rural locations was difficult to achieve, especially because some sites did not have any outdoor areas, which posed difficulties with regard to the provision of smoking areas once the smoking rooms had been closed. It also transpired that at one trust there were concerns related to the compromised safety of staff and patients leaving the perimeter to smoke in adjacent and partly unsafe public grounds and this prevented the extension of the smoke-free policy to the whole grounds.
As described elsewhere, there was considerable evidence from the questionnaires that infringements of policy regulations occurred frequently, and in more than a third of trusts were reported to occur on a daily basis. Although interview respondents confirmed that policy enforcement was a challenge, they were also generally optimistic that it could be achieved successfully in the future, alongside a change of attitude towards smoking on a societal level.
`Once the law is implemented, people will begin to appreciate. It becomes
the norm and the culture. And we support them [service users] to understand
the principles.' (Trust 1)
One respondent believed that it was a declared responsibility of mental health settings to prepare and enable individuals for compliance with smoke-free policies, since those were now in place in public places and workplaces in the community, meaning that individuals would no longer be able to smoke when and where they liked as previously.
`We needed to help people to prepare for when the whole of our society went
smoke-free, because it would prevent them expecting to smoke wherever they
wanted, whenever they wanted. Because that wouldn't be the case when they were
living around the community.' (Trust 3)
However, perseverance in policy enforcement and proactive change management involving policy promotion and support especially of staff were regarded crucial determinants of success by all those interviewed.
`It is critical that staff is empowered and encouraged and given good
guidance of how they should do it and what they should do if things are not
going right. What we have done in the revised version of the policy, we have
got an appendix which is guidance for all staff when dealing with persons
smoking in a non-smoking area. There is very little guidance out there for
this sort of thing for staff!' (Trust 2)
Cessation support for smokers
Just under 80% of questionnaire respondents reported advertising the NHS Stop Smoking Service at their sites.13 During site visits of four trusts that had responded to the survey, this information was confirmed to be available in publicly accessible waiting and reception areas. All seven interview respondents described that smoking cessation support for patients was ensured by close collaboration with the local NHS Stop Smoking Services, and six trusts offered cessation support onsite, for example through smoking cessation advisors from the primary care trust. In one trust, level 1 training in smoking cessation was about to become mandatory for nursing staff under the new policy and level 2 training was available. Six respondents described the assessment of smoking status on admission as a standard procedure, and one respondent reported that a clinical pathway for smoking cessation was pursued for those individuals who wished to receive support in smoking cessation. One interview partner reported that staff tried to find flexible and individual solutions for patient needs, and that in one case alternative therapies had been used successfully to help a smoker quit after standard support had proven repeatedly ineffective. Six of the interviewees commented that their occupational health departments offered support to staff who were attempting to stop, and two indicated that free or reduced price nicotine replacement therapy could be supplied on request. However, two interviewees mentioned that support offers were generally not taken up very well by staff.
Positive impacts of smoke-free policy
The impact of smoke-free policy implementation was explored during the interviews and was rated positive overall, despite the reported challenges described above. Reduced exposure of patients and staff to tobacco smoke as a result of closing down former smoking rooms, an enhancement of patients' motivation to stop smoking, as well as anecdotal evidence for reduced smoking prevalence in staff were named by three respondents each. Four respondents commented on the beneficial impacts the smoke-free policy had in terms of behavioural changes in patients: one respondent reported that patients were sleeping better as a result of closing smoking rooms at social gathering points where the consumption of nicotine and often caffeine had been frequent and heavy both during days and evenings. Another said that patients were reported to get up and out of their rooms earlier in the mornings, since they were no longer allowed to smoke indoors. Three respondents reported that individuals explicitly welcomed the use of newly created recreational spaces that had been provided in former smoking rooms, and that this was having a positive impact on their behaviour and sense of well-being.
`The interesting thing about this is, patients are sleeping better, they go
to bed in time. When the smoking room was available, and you couldn't sleep,
you went to smoke – now you try to sleep. And also, the smoking room
became a social area, and as soon as you were in there, you smoked. Now, we
turned the smoking room to a gym... another smoking room we use as a
relaxation room.' (Trust 1)
One respondent mentioned positive feedback from non-smoking patients who perceived the discontinued exposure to tobacco smoke on the ward as beneficial. Two respondents believed that the impact of the smoke-free policy was beneficial in terms of supporting a holistic approach towards improved physical well-being.
Although the majority of questionnaire respondents reported having used Health Development Agency guidance during policy development,12,13 some interview respondents felt that contrary to this guidance it was not feasible to extend the policy to the grounds. One participant held that it was better to assign designated smoking areas across extensive grounds and know where smokers went than not to know where to find them when they smoked covertly on the premises. Also, it was stated that the maintenance of outdoor smoking areas was crucial to avoid aggression, and necessary to facilitate successful enforcement indoors.
`There was a lot of anxiety about aggression and people's needs to smoke,
but actually the reality is, when you allow them to smoke outside – I
think that's one of the reasons why we couldn't and shouldn't go for external
grounds as well, because we had to make this work for our internal environment –
then you can manage it.' (Trust 3)
Questionnaire responses showed that 61% of trusts had communicated the policy comprehensively among staff before and during implementation, with the attribute `comprehensive' being ascribed to communication during special events or meetings, and, in addition, at least two other means of dissemination were used, for example emails, newsletters or the trust intranet. Just over 80% reported that staff were supportive of the policy to a `reasonable' extent, with the remaining percentage of respondents stating the policy was supported by staff members `very much'. More than half of respondents (52%) believed that the level of policy support by staff differed among staff groups, with nurses being most frequently identified as the least supportive group (32%).
As reported previously,13 92% of questionnaire respondents believed that the implementation of the policy had been quite or very successful. There was no notable difference in the perception of policy success between those whose policies referred to the buildings only as opposed to those comprising buildings and grounds, with the distribution of percentages in the two categories not diverging by more than 3% from those given above respectively.
Specific difficulties for mental health settings
Almost all questionnaire respondents indicated specific concerns related to smoke-free policy implementation in mental health settings, especially with regard to the high prevalence of smoking and safety issues. This finding corresponds with the results of previous work in which scepticism and anxieties with regard to the implementation of smoke-free policies in mental health settings, mostly related to increased incidences of violence and aggression, as well as fire risks, have been described.9,14,15 Interestingly, none of the interview respondents could cite examples to justify these fears, and indeed three respondents assured that all concerns had proven unfounded, with no apparent increase in aggression or violent incidences observable in the context of or after smoke-free policy implementation. The discrepancy between anticipated adverse effects and actual occurrences had been described previously in a review of smoke-free policy implementation in mental health settings.16 The finding was, however, not validated quantitatively, and was based on respondents' personal perception. In recognition of the fact that most interview respondents fulfilled a non-clinical role in their trust, it appears necessary to confirm these results with clinical staff, and in a larger sample. In this context, it would also be important to investigate further the support offered to staff and patients in dealing with smoking abstinence or cessation, as this was not a focus of this study, but will be a crucial determinant of success for smoke-free policies in practice. Furthermore, complex issues of clinical relevance, such as potential interactions between antipsychotic medication and smoking, need to be explored in greater depth.
It is notable that the majority of trusts (64%) had chosen to introduce policies that included regulations for the grounds or parts of the grounds as well. There was also evidence that different practices were applied at different sites of the same trust. Furthermore, it became clear during the interviews that the enforcement of smoke-free policy across the whole premises posed particular difficulties. According to interview respondents' accounts, these referred to specific arrangements that need to be considered for people who had been sectioned, as well as in response to varying environmental circumstances at different sites or extensive grounds surrounding the premises. Further inquiry into the feasibility and desirability of complete smoking bans at mental health sites, with a special focus on exploring site-specific challenges for patients and staff, needs to be undertaken.
Since nurses were named as the staff group least supportive of the policy, and notably fewer respondents from mental health settings (20%) than those from acute settings (40%)13 believed staff supported the smoke-free policy `very much', it would appear useful to explore the nature of and reasons for this perceived lack of support. It will be important to secure staff support for smoke-free policies in mental health settings in the future. In order to do this, it will be important to acknowledge the difficulties faced by staff in everyday practice and comprehensive support offered.
As indicated by interview respondents, further guidance for mental health settings will probably be needed to ensure implementation is not undermined.
Impacts of smoke-free policy
Almost all questionnaire respondents believed that the implementation of their smoke-free policy had been quite or very successful, although this perception appears somewhat counter-intuitive when considering the high level of policy infringements, the breadth of exemptions that were reported to be granted and the fact that nurses who are central in dealing with the policy were described as least supportive. This ambivalence was, however, reflected during the interviews, with respondents clearly feeling that the policy was overall a success despite its challenges. Beneficial impacts cited mainly related to reduced exposure to tobacco smoke, enhanced motivation of patients to stop smoking, and reduced smoking prevalence among staff were identified, all supporting findings of previous research8,17,18 and indicating that smoke-free policies can constitute an important component of health promotion. Interestingly, several interview respondents could report behavioural changes following smoke-free policy implementation in patients, for example related to sleeping habits. Changes were also apparent where the closure of smoking rooms had resulted in the creation of health-promoting recreational spaces, as had been reported in an earlier study.10 Further qualitative and quantitative inquiry into the effects of smoke-free policy in relation to a change of the ward environment and potential impact on individuals' mental health condition and well-being would therefore be useful to elucidate the potential for health promotion in mental health settings.
General study limitations referring to the legal and political requirements relating to smoke-free policies, and to the fact that 21% of potential participants did not respond to the questionnaire, have been detailed elsewhere.13 The choice of one study participant per trust constitutes a further source of bias because the perspective was generally restricted to that of a non-clinical executive, and the information provided likely to refer predominantly to one site of the respective trust, whereas all trusts comprised multiple sites. The inclusion of perspectives from clinical staff on different trust sites would have complemented the data usefully, but identifying appropriate contact persons in remote sites without local knowledge is extremely difficult. Since the interview sample was drawn from the 46% of questionnaire respondents who agreed to participate in an interview, interview data may be affected by self-selection bias. Generally, interviews and site visits were conducted in small samples, so these findings need to be regarded as preliminary and should be verified in further research. Site visits would have benefited from obtaining permission to access non-public trust areas for detailed observation.
This study shows that considerable efforts to implement smoke-free policies in English mental health in-patient settings have been undertaken, and reveal that overall the outcome is rated positive. However, challenges are widely perceived and some concerns specific to mental health settings have been identified. These need to be explored further and addressed adequately to support trusts in complying with the Health Act from July 2008 and in maximising the benefit of the law.
- Received February 8, 2008.
- Revision received September 29, 2008.
- Accepted October 6, 2008.
- © 2009 Royal College of Psychiatrists