Abstract
This paper serves as a review of the evidence for the feasibility of smoking bans in psychiatric inpatient settings. A brief summary of the literature on smoking and mental illness and a rationale for investigating this issue as a major public health concern provide the context for the timeliness of this review. The Australian experience is reflective of similar concerns internationally [1], [2].
Smoking as a public health problem
Links between smoking and cancer of the lung were first confirmed by Doll and Hill [3], [4]. The Royal College of Physicians published the first major authoritative report on smoking and health in 1962 [5], being the forerunner to many other major reports, such as that of the US Surgeon General in 1964 [6]. Since that time, the links between smoking and disease have been well established, with more than 57 000 scientific articles published on this subject [7]. Both the Royal College of Physicians and the US Surgeon General have been responsible for several of these reports [5], [6],[8–13]. Since then, the evidence for smoking as a serious public health concern has been growing. Tobacco smoking accounts for 3–5 million deaths worldwide each year, with this figure predicted to reach 10 million per year in the decade 2020–2030 [14]. Globally, tobacco is the leading risk factor for disease burden [15]. Indirect exposure to smoking as a result of environmental tobacco smoke or passive smoking has also been confirmed worldwide as a significant public health problem [16–19].
Comorbid nicotine dependence and mental illness
Smoking prevalence is among the highest for people with mental illness; up to 88% for those with mental illness compared to approximately 25% for the general population. Research has also clearly established that mentally ill smokers tend to smoke more heavily, for more years and favour higher tar cigarettes than the general population [20–22]. Using data from the National Survey of Mental Health and Wellbeing of Adults in 1997, Jorm [23] found this association to be particularly prominent in the 18- to 39-year-old age group. Despite the vast body of literature and research on cigarette smoking, the majority of research has concluded that quit rates for people with a concurrent mental illness continue to be extremely low [20], [21],[24–27]. The high prevalence of smoking among all people with a mental illness is a concerning public health problem. Links between smoking and higher premature death rates from all major physical health conditions have been noted for this group when compared to the general population [28–30]. The presence of fewer health-promoting behaviours and poorer nutrition, generally for people with mental illness, has also been proposed to help explain their greater risk of premature death [31].
Reviews of the existing research on smoking and mental illness have found significant comorbidity with several pharmacological and psychosocial reasons for this comorbidity proposed [1], [32], [33]. Smokers with schizophrenia are thought to use cigarettes to self-medicate the effects of negative symptoms of their illness [27], [34], [35]. Smoking has also been reputed to have antidepressant effects in people suffering from unipolar depression with smoking cessation attempts being causally implicated in the relapse of these people's depression [36], [37]. Research has also shown that smoking relapse is more likely in the presence of negative mood states [38]. Nicotine's role in regulating a dysfunctional dopamine system, by augmenting dopamine release, has been proposed as the mechanism involved in smoking dependence for people suffering from schizophrenia and depression [36]. More generally, central nervous system mesolimbic dopaminergic pathway activity has been found to be especially important in mediating reward in nicotine dependence [39], [40]. Smoking has also been shown to mitigate the side-effects of neuroleptic medications that are widely used by psychiatric populations, to treat their mental illness. One such side-effect, neuroleptic-induced parkinsonism, has been increasingly found to be less common in smokers [20], [27], [41], [42]. Recent biological in vivo research with non-psychiatric populations has confirmed that smoking and the development of dependence are associated with increased dopamine activity in the basal ganglia and that smokers have special sensitivity to presynaptic dopaminergic activation by nicotine [43]. The role of nicotine in improving cognitive function has also been proposed, with mentally ill smokers reporting that smoking helps to overcome deficits in attention, concentration, memory and cognitive functions generally. Nicotine has been shown to improve sensory gating so that smoking alleviates sensory information processing difficulties. Auditory sensory gating deficits are found in more than 75% of people with schizophrenia and these deficits are temporarily normalized by smoking for these people. However, it is unclear whether nicotine has direct positive effects on cognitive function in smokers or whether it plays a role in reversing cognitive deficits [44–46]. What may be of greater relevance is the notion that, once smoking and addiction become established, smokerswith mental illness may find quitting more difficult because of a range of psychosocial reasons such as impairments in social and cognitive functioning [47], and problems associated with anxiety, medication side-effects, motivation and lack of other coping resources [48]. Therefore, cessation programs that rely on the transtheoretical model, with its emphasis on motivation levels and readiness to change, may not be appropriate for this group of smokers [49], [50]. Smoking has also been proposed to have a protective effect against dementia, but this has not been confirmed in a report reviewing the evidence [51]. The existential, social and cultural influence of psychiatric settings and mental illness on smoking rates for staff and patients has been explored elsewhere [33], [50], [52], [53].
Smoking bans in psychiatric settings
The culture of smoking in psychiatric settings is perceived to be an entrenched process that has been central to the history of mental institutions over the past three centuries with the development of asylums and their evolution into our current psychiatric inpatient facilities. Tobacco rations were an assumed part of day-to-day life in many such institutions [54]. The idea of imposing smoking bans in psychiatric settings is thought to be a recent phenomenon. However, there is evidence for much antitobacco sentiment, for example, in the 1800s in the USA. In the 1830s and 1840s Samuel B Woodward, the Superintendent of the Worcester State Hospital in Massachusetts, wrote vast commentaries raising the harms of smoking [55]. In 1848, an article in the American Journal of Insanity by Dr Pliny Earle, the Superintendent of the Bloomingdale asylum in New York, concluded that ‘smoking is considered so deleterious that in most of the well-conducted establishments for the insane in this country, its use among the patients is prohibited. At this institution it is not permitted, excepting in a few cases, in small quantities, by patients who have resided here many years’ [56].
The British College of Physicians and US Surgeon General reports of the 1950s and 1960s highlighted the physical harms of smoking and triggered a new wave of concern. These reports eventually influenced and prompted a number of US psychiatric institutions to introduce smoking bans from the late 1980s and early 1990s. In 1992, the US Joint Commission on Accreditation of Healthcare Organizations declared that hospital buildings must adopt the goal of eventually becoming smoke-free. The following reviewof 26 studies documents the experiences of these and later psychiatric studies of settings, where smoking bans were introduced. The review would seem timely because of the recent proliferation of research in this area and increasing activity in and demands from the practice field for clear policy to guide solutions to this dilemma. All these studies are useful for their articulation of the processes they followed in order to achieve smoking bans and the lessons they learned along the way.
Method
This review builds on an earlier review by Patten et al. [57]. The search strategy used for the review of research on smoking bans in psychiatric settings involved a general electronic database search of Pubmed using the terms (tobacco use disorder OR smoking OR smoking cessation OR cigarette∗) and (hospitals∗ AND mental disorders OR psychiatric hospitals OR psychiatric department, hospital). The search was restricted to English language and included any sources from 1970 onwards. One hundred and eighty records were retrieved. PsychINFO was also searched using the terms nicotine or smoking, smoking-cessation, tobacco-smoking, psychiatric hospitals, or psychiatric units. This search was also restricted to English language and included any sources from 1970 onwards. Thirty-six records were retrieved. Reference lists used in each relevant research paper were also examined as well as existing policy documents on the topic of smoking and mental illness generally. The main author also routinely checked a broad range of journal publications by means of the Elsevier Science Contents Direct electronic alert system. As research in this area is limited, all known studies were included. 33]. Therefore, the results of those studies where a total ban is genuinely applied to the settings are also defined.
Studies on smoking bans in psychiatric treatment units
Summary of key findings from the 26 reviewed studies
Discussion
Overall, the findings of these studies are mixed. Unintended negative consequences of change are evident in each study presented. However, staff generally anticipated more smoking-related problems than actually occurred. Some researchers stated that few transition problems were experienced by patients and staff, while other studies clearly present some concerning findings.
Of greatest significance was that most studies found that therewas no increase in aggression, discharge against medical advice or increased use of as-needed medication following the ban. This was the case for approximately 75% of all study sites regardless of the type of ban imposed and in 90% of sites that imposed total bans. Of the two study sites that reported an increase in these problems with the imposition of a total ban, the first described four case studies of highly disturbed patients who were detained and unable to enter the grounds to smoke. This study also noted problems with no administrative process to provide consistent enforcement of the ban, suggesting fragmentation may account for these problems [58]. This need for consistency of approach by staff, ranging from management to clinical staff support, was noted by several studies to be important for success. The concerns for staff morale and anxiety levels as part of a change process and the destructive effects of not having a consistent approach were noted in several of these studies and elsewhere [33]. There was also no mention of staff education about differentiating between psychotic symptoms of distress and nicotine withdrawal symptoms for patients at either of these sites, which may also have contributed to this negative result [52], [58]. The impact of fragmentation and inconsistent application of bans across the patient population tended to cause more harm and disruption as experienced by studies that tried to impose selective bans. Where restrictions are graduated over time, they can have the unintended consequence of focusing on the negotiation of smoking privileges, increasing the value of cigarettes as a tool for exchange and therefore heightening the potential for conflict [33], [50], [59]. This is exactly what 11 of the 26 studies found (key findings 6 and 7).
When questioned before the implementation of bans, most staff, particularly nursing staff, predicted more adverse effects than actually occurred. However, they developed amuch more positive viewpost-ban. Thiswas noted in approximately 55% of studies overall and in 70% of studies where total bans were imposed. The initial fears of nursing staff can be attributed to staff in this profession playing the most significant role in providing direct care to patients, more so than other disciplines within psychiatric settings. Therefore, nurses are arguably more likely than others to be assaulted by agitated patients and to develop extremely close nurturing roles with patients and identify strongly with patient distress, nicotine withdrawal being one of these. There is also a vast literature on the high rates of smoking by psychiatric nurses, compared to other nurses and other health professionals [60–63]. When smoking bans have been imposed, the rate of smoking by staff has been shown to decline with many staff taking the opportunity to quit once bans are imposed [64], [65].
The design of many of these studies appears to be weighted disproportionately at researching the impacts of bans on the staff and the institution itself, rather than on the impacts on patient wellbeing. This is evidenced by the lack of consideration many of the studies give to patient quit rates and relapse rates. The impact of bans on staff quit rates is likewise scantily covered and omitted by most of these studies. Initiation into smoking or relapse to smoking, as a result of a strong smoking culture in inpatient settings, has been acknowledged as a significant problem for people who are admitted to these settings [33]. It would therefore seem of great interest to measure what the impact of imposing bans would be. Clear policies and collaborative partnerships between hospitals and community services are needed to provide continuous and consistent pathways of care and support. This is essential if the gains achieved in inpatient settings where bans are imposed are to continue in the community.
Central to the notion of change is the need to understand why change is often perceived as so difficult to achieve. Schon's [66] concept of ‘dynamic conservatism’ is a useful one and is supported by Ogburn [67]. They suggest that organizations are resistant to change and that staff and patients tend to use existing forms of behaviour management, out of habit, rather than create new ones. The accepted use of cigarettes by staff to manage patients in mental health settings before imposing bans may have acted as the mechanism for many of the rules of interaction, and procedures and actions taken in the settings. Once a ban is imposed, many of these rules need to be renegotiated. This can be a difficult transition for all concerned, dependent on the consistency of and commitment to the new approach, provision of education and other supports to both staff and patients, and other potential factors that influence cultural change in the setting. Six of the reviewed studies noted that staff saw the bans as an opportunity to learn new clinical skills.
Conclusions
This review is based on research from three countries (US, Canada and Australia) and most of it is from the US. This may limit generalizability of findings to countries which are culturally similar to these. In general, the findings show that a number of measures would need to be considered in order to introduce effective smoking bans.
The over-reliance by nursing staff on smoking to assist with the clinical management of patientswould need to be addressed. Helping nursing staff to find alternative options is seen as essential. The use of nicotine replacement therapy (NRT) by patients as part of imposing the ban is shown to improve success. Extensive consultation and collaboration, coordination of efforts across the disciplines, provision of alternative activities, dietary changes, clear protocols and family support for the bans would need to occur. More effective measures to accommodate patients who are unable to tolerate abrupt abstinence would be needed. Greater awareness of the ban before admissionwould be useful. This would involve coordination and partnership across the mental health sector between community and inpatient services. Greater support for and education of direct-care staff on distinguishingmental illness symptoms from nicotine withdrawal symptoms is seen as vital. This would require support at all levels, from direct care of patients to hospital administration and policy. A preparation period, before the ban, involved community agencies and groups and inpatient staff involving education and advertising of the impending ban to patients is also proposed. Where staff are banned from smoking at work, alternative supports would need to be developed to assist staff to manage their own stress levels and to clinically manage patients. Patients may interpret restrictions as a further source of powerlessness and control by others, with implications for staff morale as agents of further social control. This would need to be addressed with open and equitable consultation with all parties. Trade and standover for cigaretteswithin the grounds of the hospital may increase, with potential for such activities to increasingly spread beyond the grounds to nearby shops, houses and the community generally. A planned transition to the ban with widespread consultation and implementation of strategies would be needed. Black market use and sale of tobacco within mental health settings may increase. Use of other drugs may increase. Nicotine interacts with antipsychotic drug metabolism so that patients tend to need more medication when they smoke and less when they quit smoking. There is also a high expectation that many patients would return to smoking upon discharge from hospital. Therefore, patients who have been banned from smoking while in inpatient settings, who then return to smoking upon discharge, may need their medication reviewed to account for this change. Community mental health teams would need to be aware of this as part of improved coordination of follow-up. Given that many patients returned to smoking postdischarge, it is clear that bans alone were not effective in assisting people to quit in the longer term. Imposing bans in inpatient settings is seen as only part of a much larger strategy needed to overcome the high rates of smoking among mental health populations, generally. More coordinated efforts would be needed between hospital and community staff to help patients who wish to stay quit as part of discharge planning. Mental health services would need to develop clearer policies with regard to smoking and occupational health and safety concerns for staff and patients as part of the process of imposing bans and maintaining them. This would include clearer clinical and ethical guidelines that address the issue of distress and withdrawal, patient autonomy and legal aspects of imposing a ban [40].
This review has shown that the introduction of smoking bans in psychiatric inpatient settings is possible but would need to be a clearly and carefully planned process involving all parties affected by the bans. Consistency, coordination and full clinical and administrative support for smoking bans are seen as essential to their successful implementation.
