Exploration of in-patient attitudes towards smoking within a large mental health trust

AIMS AND METHOD

To explore the smoking habits of in-patients on psychiatricwards, their beliefs about the effects of smoking on health,and their attitudes towards hospital and government smokingpolicies. Face-to-face interviews with 135 in-patients wereconducted.

RESULTS

A total of 54.1% of participants smoked. Smoking was less prevalentin those aged 65 years and older (P<0.001). Non-smokerswere more likely to believe smoking to be harmful to health(P=0.002). Overall, 71.1% of the participants favoured theexisting smoking policy, with only 3.0% wanting a completeban on smoking and 54.1% agreeing with banning smoking in publicplaces.

CLINICAL IMPLICATIONS

Further studies are warranted into the views of in-patientselsewhere and to see whether attitudes change as trusts tightentheir smoking policies. Outdoor smoking areas may need to beconsidered, although in practice this may not be possible inall trusts.

Introduction

TOP

Introduction

On 1 July 2007 all enclosed or substantially enclosed publicplaces and workplaces became smoke free as required by TheHealth Act 2006. A temporary exemption for mental health unitsends on 1 July 2008. Until then they may have a designatedsmoking room meeting specified requirements.

A very small number of studies exist on attitudes of staff towardssmoking in psychiatric hospitals in the UK (Tarbuck, 1996; Stubbs et al, 2004), with even less information regardingpatients’ views. With impending government legislationand potential plans within Mersey Care NHS Trust to restructurethe in-patient units, this was an ideal time to investigatethis topic.

Mersey Care NHS Trust provides mental health services for thepopulation of Liverpool, Sefton and Kirkby. At the time wesurveyed its wards, the Trust had a general non-smoking policy.This entailed one or two smoking rooms on each ward with allother enclosed areas being non-smoking.

Our study aimed to explore the attitudes of in-patients acrossthe General Adult and Old Age Directorates of the Trust towardsthe smoking policy and towards the perceived effects of smokingon health and well-being. We also aimed to examine any changesin smoking habits during hospital admission.

Method

TOP

Method

A questionnaire was constructed following a review of literatureon this topic (Senior, 1982; Stubbs et al, 2004; Willemsen et al, 2004)and based on the previous discussions with patients and ward staff. Structured and open-ended questions were included togenerate both quantitative and qualitative data. The studyprotocol was then submitted to the Liverpool Research EthicsCommittee Coordinator.

During April and May 2006, face-to-face interviews were conductedwith in-patients accepted for the study on ten general adultand three functional old age wards in the Trust hospitals.Patients were excluded from participation if their conditionwas too unstable. Written consent from each patient was obtainedprior to interview.

The results were analysed using SPSS version 14.0 for Windows.Differences between smokers and non-smokers, under 65-year-oldsand over 65-year-olds, and those detained and informal weretested with the Pearson chi-squared and Fisher’s Exacttests, both two-tailed. Since there was a higher number ofsmokers among younger patients ( ${chi}$ ²=14.28, P<0.001), resultspertaining to age were standardised according to current smokinghabits. Ex-smokers were reclassified as non-smokers to reduce the number of analyses.

Results

TOP

Results

Patient characteristics
Out of 243 in-patients on the 13 wards, 135 were successfullyinterviewed (response rate 55.6%; 52.6% men, 47.4% women).The mean age of interviewees was 49.7 years (s.d.=16.7, range18-86), with 76.3% aged less than 65 years. A total of 68.1%of the participants were in informal care and 15.6% had beenin hospital for at least 6 months.

Smoking status
The overall percentage of current smokers was 54.1%, with 54.8%smoking prior to admission. No differences were found withrespect to gender, Mental Health Act status or length of stayin hospital. However, more of those younger than 65 years oldsmoked than those aged 65 years and over (63.1% v. 25.0%; ${chi}$ ²=14.28,P<0.001). Within the smoking group, 59.7% smoked at least20 cigarettes per day (Table 1).

View this table:
[in this window]
[in a new window]

Table 1. Comparison of questionnaire responses by gender, smoking status, Mental Health Act status, age or the length of stay in hospital

Changes in smoking habits
Although the reported overall rate of smoking remained fairlyconstant between pre-admission and the time of interviewing,14.1% of the participants reported an increase and 23% a decreaseof the habit. Two people reported re-starting smoking, tworeported smoking for the first time in their lives and tworeported both increasing and decreasing of smoking since admission.

To give reasons for increasing smoking, participants could choosethose provided in the questionnaire - ‘boredom’(n=18), ‘stress’ (n=12) and ‘to socialise’ (n=6) - and cite their own, for example ‘enjoying smoking’(n=1), ‘craving more when unwell’ (n=1) and ‘theenvironment’ (n=1).

Reasons for decreasing smoking were the provided options: ‘health concerns’ (n=17), ‘financial reasons’ (n=7)and ‘advice from staff’ (n=1), and other reasonslike ‘smoking restrictions on the ward’ (n=7), ‘difficultyobtaining cigarettes’ (n=3), ‘shouldn’t smokein hospital’ (n=2), ‘want to give up smoking’(n=1), ‘pressure from relatives’ (n=1) and ‘lessstressed in hospital’ (n=1).

Attitudes towards effects of smoking on health
The number of participants who believed smoking to be harmfulto health was 92.6%, with smokers less likely to believe thisthan non-smokers (86% v.100%, Fisher’s=9.17, P=0.002).As much as 84.4% of the participants thought passive smokingto be harmful to health, with no difference between smokersand non-smokers, and 2.2% were not sure.

Attitudes towards general non-smoking policy
Fifty-seven participants (42.2%) reported difficulties withthe current general non-smoking policy. These results are shownin Table 2. Those non-smokers younger than 65 years old weremore likely to report difficulties than non-smokers aged 65years and over (Fisher’s=9.11, P=0.003).

View this table:
[in this window]
[in a new window]

Table 2. Difficulties experienced with the current smoking policy

Preferred smoking policy within the Trust
Only 3.0% chose complete ban inside and on premises as theirpreferred smoking policy, 14.1% supported complete ban insideonly, 71.1% supported a general non-smoking policy with designatedsmoking areas, 7.4% a general smoking policy with non-smokingareas and 4.4% would like no restrictions on smoking.

Attitudes towards government proposals
As much as 54.1% of participants in the study agreed with thegovernment proposals to make all public places non-smoking.Non-smokers were more likely to agree than smokers (77% v.34%, ${chi}$ ²=25.2, P<0.001) and informal patients were more likelyto agree than detained patients (61% v.40%, ${chi}$ ²=5.37, P=0.020).

Appendices (see online data supplements)

TOP

Appendices (see online data...

Outline of the qualitative data generated from additional comments(Table DS1).
The complete data on the comparison of questionnaireresponses(Table DS2).

Discussion

TOP

Discussion

The higher smoking rates among younger people in this studywere also found by the General Household Survey in the generalpopulation (Office for National Statistics, 2005a). That moresmokers reported a decrease in smoking (23.0%) than an increase(14.1%) contrasted with our expectations, based on anecdotalevidence, that smokers might smoke more in hospital. Hence,we feel that the assumption that it would be particularly difficultto target non-smoking policies at psychiatric in-patients maynot be justified.

While a high proportion (92.6%) of the patients in this studyrecognised smoking to be harmful to health, we did not ascertainwhether they appreciated the extent of the harm caused. In2005, only 48% of the population was aware that smoking isresponsible for the most premature deaths in the UK (Omnibus Survey; Office for National Statistics, 2005b). The high percentageof patients (84.4%) in our study recognising passive smokingas being harmful to health is compatible with the Omnibus Surveyfindings in 2005 when over 80% of participants thought second-handsmoking would increase a non-smoking adult’s risk of lung cancer, bronchitis and asthma.

The finding that non-smokers younger than 65 years old reportedmore difficulties with separate smoking areas on the wardsthan non-smokers aged 65 and over may result from the greaterproportion of smokers in the under-65-group. Alternative explanationsmay be different ward layouts, greater tolerance towards smokingamong older people who grew up in a more smoking-permissivesociety and the fact that older people have been found to bemore reluctant to express dissatisfaction with hospital policies (Calnan et al, 2003).

Although non-smokers and smokers were equally likely to haveproblems with the existing policy, the majority of complaints(80.8%) were anti-smoking in nature. Nevertheless, despitethese difficulties and knowledge of the harmful effects ofsmoking, the current general non-smoking policy was by far themost popular (71.1%). This may indicate that in-patients onpsychiatric wards are generally tolerant towards others. Alternatively,it may reflect the problem of social desirability, which commonlyappears in surveys. In contrast with our study, a large Dutchsurvey of staff and patients across various psychiatric settingsfound that 45% of patients preferred a general non-smokingpolicy (Willemsen et al, 2004). However, only 8% preferreda complete smoking ban which compares with 11.8% in our studyfavouring the more restrictive policies.

There was a large difference observed in our study between thosewanting a total smoking ban inside hospital buildings (14.1%)and those supporting the government proposals of banning smokingin public places (54.1%). This may reflect views that smokingpolicies should be more lenient in psychiatric units, as highlightedby the comments of eight patients. We found a considerablylower percentage of participants agreeing with the proposedban on smoking in public places (54.1%) than was seen in thepublic consultation on the Health Bill (89.6%; Department of Health, 2005).Since our data shows that non-smokers were more likely to agreewith this aspect of the legislation than smokers, this may be one explanation. Alternatively, our proactive method (face-to-faceinterviews) might have encouraged a wider variety of respondents.

There were some limitations to this study, namely volunteerbias, recall bias and slight environmental differences betweenwards. The number of hypothesis tests would have increasedthe likelihood of chance findings. Conversely, the small numbersin some groups may have meant insufficient power to detectadditional significant differences. Lastly, ex-smokers were re-classified as non-smokers although these two groups mayhave had different views.

Our study reinforces previous findings that smoking is moreprevalent in patients on psychiatric wards than in generalpopulations (de Leon et al, 1995; Meltzer et al, 1996; Lasser et al, 2000).

As the majority of in-patients, both smokers and non-smokers,felt there should be smoking areas within psychiatric units,can any appropriate compromise be reached there in view ofthe current legislation? One suggestion would be to ban smokinginside buildings in line with government legislation, but tohave unenclosed smoking shelters outside the buildings, awayfrom potential routes of entry of smoke into buildings. However,this may mean staff escorts being exposed to second-hand smoke.Furthermore, only patients well enough to leave the ward wouldbe able to smoke. Another option could be outdoor smoking areasleading off from and visible from wards, but the problem ofsmoke infiltrating buildings remains and could necessitate theredesigning of buildings in many trusts.

It would be interesting to know if these results are mirroredelsewhere in the country and whether patients’ viewsare changing following the implementation of tighter smokingpolicies within NHS trusts. It would also be worth evaluatingthe level of compliance with such policies.

Acknowledgments

We thank Dr Steven Lane, Senior Research Assistant at the Centrefor Medical Statistics, University of Liverpool, for his helpwith analysis of the results, and Dr Robert Higgo, ConsultantPsychiatrist, for his comments on the manuscript.

References

TOP