- Email: [email protected]
Why nurses smoke: a review of the literature
International Journal of Nursing Studies 37 (2000) 173±181
www.elsevier.com/locate/ijnurstu
Why nurses smoke: a review of the literature Kathy Rowe a,*, Jill Macleod Clark b a
The School of Nursing and Midwifery, The Queen's University of Belfast, University Road, Belfast, Northern Ireland, UK b University of Southampton, Southampton, UK Received 2 March 1999; accepted 6 June 1999
Abstract The smoking behaviour of nurses has been widely debated in the context of their professional role and responsibilities. There has been much speculation about why nurses smoke and possible explanations include a stressful nursing environment, peer pressure and socio economic status and education. This paper provides an overview of the literature which oers insights into the reasons why nurses smoke and compares the ®ndings from this literature with those studies examining the smoking behaviour of women in general and young women in particular. This review reveals that many students take up smoking before commencing their training and the factors which in¯uence nurses smoking are similar to those that in¯uence similar groups of females in the general population. # 2000 Published by Elsevier Science Ltd. All rights reserved. Keywords: Nurses smoking; Women smoking; Student smoking; Adolescent smoking
1. Introduction Over the years there has been an ongoing debate about the implications of the smoking behaviour of nurses in relation to their own health and their health promotion role. (Faulkner and Ward, 1983; Macleod Clark et al., 1987; Rowe, 1998). Empirical data about the smoking behaviour of nurses is limited with many studies being out of date or based on small samples (Adriaanse et al., 1991; Rowe, 1998). There seems to be a general consensus from the literature that the incidence of smoking amongst nurses equates with that of women in the population in general, and is around
* Corresponding author. Tel.: +44-1232-273997; fax: +441232-319356. E-mail address: [email protected] (K. Rowe).
28% (Oce for National Statistics, 1996). This rate in nursing is in¯uenced by status and professional role, with those in more senior managerial positions smoking less than student nurses, enrolled nurses, new sta nurses and nursing auxiliaries (Becker et al., 1986; Feldman and Richard, 1986). This again equates with variations in smoking rates in the general female population which relate to occupational status (Oce for National Statistics, 1996). A number of explanations have been oered in an attempt to establish why nurses smoke. The three main themes that emerge from a review of the literature are those of stress, peer and social in¯uence and socio economic status and education. The review which is presented under these three headings was undertaken using a number of computer data bases but with particular scrutiny of CINAHL 1982±1997 and Medline 1992±1997.
0020-7489/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII:S 0 0 2 0 - 7 4 8 9 ( 9 9 ) 0 0 0 6 0 - 7
174
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
2. The in¯uence of stress on smoking amongst nurses Many studies of nurses' smoking behaviour have attempted to demonstrate that smoking is a coping mechanism against stress, caused by the nursing environment (Leathar, 1980; Tagliacozzo and Vaughn, 1982; Spencer, 1984; Booth and Faulkner, 1986; Elkind, 1988; Cinelli and Glover, 1988; Carmichael and Cockcroft, 1990; Charlton et al., 1997). While a number of these studies support this theory, con¯icting ®ndings are also reported. The studies are presented in chronological order and re¯ect the smoking behaviour of student nurses1 pupil nurses2 and quali®ed nurses. Leathar (1980) surveyed the smoking behaviour of 93 nurses of all grades in order to investigate the reasons why they smoked. He employed in-depth interviews, each lasting at least 1 h and focused on understanding nurses' smoking in relation to stress speci®c to nursing. Participants were encouraged to identify particular areas of diculty, such as death, night duty, a rigid authority structure and responsibility problems. Leathar found a crucial time appeared to be during the ®rst 18 months of training when some nurses started smoking and others who already smoked increased their levels of smoking. A number of reasons were given for smoking, such as not knowing anyone and having to make friends, having to cope with more responsibility than one should for her status particularly on night duty and the general lack of support given by trained sta. In order to cope with the continual uncertainty, frustration and lack of support, informal friendship groups were formed and cigarettes were used as a symbol of friendship and mutual trust. While the ®ndings suggest that student nurses smoked in order to cope with stress incurred in the nursing environment, the report does not identify how many nurses were in fact smoking prior to commencing nursing. Moreover, the study does not probe to any extent the power of peer friendship groups in initiating and maintaining smoking. Tagliacozzo and Vaughn (1982) surveyed 933 registered nurses in seven hospital departments to establish whether there was an association between a stressful nursing environment and nurses' smoking behaviour. They reported that 19.9% of nurses smoked and that nurses who perceived their job as stressful were more likely to smoke than nurses who did not. It is not 1 Student nurse: applies to a student who is enrolled on a 3year programme for the professional quali®cation of registered nurse. 2 Pupil nurse: applies to a student who is enrolled on a 2year programme for the professional quali®cation of enrolled nurse. This programme is no longer available in the UK.
reported however, how many nurses were smoking prior to their entry to nursing. The ®ndings indicated that nurses most likely to perceive the physical and emotional demands of the job as stressful, were under 29 years and single. The authors acknowledge the need for further research to explore whether smoking amongst younger nurses is a mechanism for coping, or whether younger nurses who smoke are simply more likely than older nurses to perceive work situations as stressful. However, given the limited response rate (49.3%) and the use of an ambiguous de®nition of smoker the ®ndings should be interpreted with some caution. Spencer's (1984) survey of the smoking behaviour of 1577 nurses in two health authorities in England and Wales also attempted to determine whether smoking was related to stress in the nursing environment. The sample comprised all types of quali®ed nurses, student nurses and nursing assistants. The results revealed that although nurses did ®nd particular parts of their job stressful, most coped well with this stress. Therefore the idea that either the feelings of stress associated with particular nursing tasks, or the ability to cope with the stress, is connected to cigarette smoking was not supported by his evidence. In fact the ®ndings identify that more non-smokers (55.5%) felt stressed by patients dying, than did smokers (46.8%). In all other activities the amount of stress felt by smokers and non smokers was very similar, thus the view that nursing, cigarette smoking and stress are connected, may not be supported. Booth and Faulkner (1986) surveyed the smoking behaviour of 563 student and pupil nurses and found that 36.4% were smoking. Moreover 73% of the current smokers claimed to have been smoking for longer than three years. Thus the researchers concluded that since 95% of the sample were still in training, most were smoking prior to coming to nursing. In order to determine some of the factors associated with smoking, they moved beyond the questionnaire data and conducted in-depth interviews with 10% of the nurses who smoked. This data revealed that over half of the smokers in the sample perceived worry, stress or tension as contributing to their smoking behaviour. Despite this, it is interesting to note that very few (4.4%) of current smokers reported speci®c stressful nursing situations as being associated with increased smoking. In fact only nine nurses mentioned nursing at all and four of these appeared to interweave nursing with outside factors in their lives. These ®ndings are similar to those elicited by Elkind (1988) and Rowe (1998). They add weight to the view that while nurses use cigarettes as a coping strategy for the stressful situations they encounter in their everyday lives, their smoking behaviour should be considered in a wider context
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
than that of stress generated by the nursing environment alone. Elkind (1988) surveyed 43 undergraduate student teachers3 and 69 learner nurses4, during their ®rst year of training. The aim of the study was to explore the smoking behaviour of student nurses and teachers at entry to training, and to determine the factors which underlie the initiation of smoking. The design of the study included a baseline questionnaire administered at commencement of training, followed by an interview during the induction term. Three further questionnaires were distributed during the ®rst year of training at 5 months, 8±10 months and 12±15 months after entry to their course, followed by a second interview. The ®ndings revealed that 33% of student nurses were smokers at entry to their nursing programme, compared with 16% of student teachers. This again supports the view that a large number of nurses do commence their smoking habit prior to commencing their nursing career. Pre-eminent among the student nurses and student teachers responses (smokers and non-smokers alike), was the idea that smoking relieved stress (52%). The concept of stress was further explored during the second interview and it appeared that both nurses and student teachers alike experienced a signi®cantly greater degree of stress when in the practical environment than in the academic environment. While Elkind concludes that stress among student nurses is an issue, it would seem that patterns of smoking behaviour may be established at entry to training and therefore the role of stress should not be interpreted over simplistically. Cinelli and Glover (1988) surveyed the smoking behaviour of 299 registered nurses in Pennsylvania utilising an adaptation of the Job-related Stress and Smoking Behaviour Questionnaire developed by Tagliacozzo and Vaughn (1982). The questionnaire which was revised to include other sections such as career history, smoking history, personal data and nurse job stress index, yielded a response rate of 49%. The purpose of the study was to examine the level of professional nurse training of nurses who smoked, the aect of social factors on the initiation and reinforcement of smoking in the workplace and home, and the impact of occupational stress on smoking behaviour. The authors reported that occupational stress was not signi®cantly related to the smoking behaviour of nurses, and the strongest predictor for smoking amongst nurses was in fact the number of friends who 3 Student teacher: refers to a student enrolled on a programme for the professional quali®cation of degree in education. 4 Learner nurses: refers to students enrolled on both registered nurse and enrolled nurse programmes.
175
smoked. A current smoker was de®ned as an individual who smoked more than one tobacco cigarette a day for a period of three months or longer during the past year. This is an unusual de®nition and no explanation is given as to why individuals who smoked one cigarette a day were excluded from the study. Again these ®ndings should be treated with some caution given the poor response rate. Carmichael and Cockcroft (1990) used a questionnaire to examine the smoking behaviour of 368 student nurses and found that 43% of the sample were smokers. They further reported that 82% of those who were smoking were smoking on entry to nursing. This once more highlights the importance of exploring the reasons why nurses smoke within a much wider context than the nursing environment. While Carmichael and Cockcroft reported that stress, greater opportunity and peer group behaviour in¯uenced 18% of student nurses to commence smoking and in¯uenced a further 18% to increase the number they smoked, they did not explore the reasons why the vast majority (82%) were smoking before coming to nursing. A stressful nursing environment cannot therefore be implicated as a major reason why nurses smoke. Charlton et al. (1997) more recently explored the smoking behaviour of 90 student nurses at classes in one university. A questionnaire approach was used and a response rate of 72.2% was achieved. The ®ndings revealed that of those surveyed, 28% were current smokers. The researchers however, acknowledge that during distribution of the questionnaire, the absence rate from class was high due to a ¯u epidemic. It is therefore possible that this percentage may not re¯ect an accurate picture of the extent of student nurses smoking in this university. Although this is a small scale study, the ®ndings are similar to that of previous studies, in that the majority of smoking students were smoking prior to coming to nursing. In fact 64% had tried their ®rst cigarette before the age of 16 years and the rest between the ages of 16 and 18 years. Moreover, levels of stress were not signi®cantly associated with smoking behaviour. This once again, suggests that the assumption about stress and smoking amongst nurses needs to be revisited, particularly given the numbers of students who bring their smoking to nursing. In general, studies which have examined the reasons why nurses smoke do seem to identify that the nursing environment is potentially stressful but they fail to comprehensively demonstrate a clear link between stress in the workplace and smoking. Findings from some of the studies are compromised by methodological weaknesses including low response rates and lack of detail to de®nition of a smoker. These researchers have attempted to examine the relationship between stress in the workplace and smoking and failed to look
176
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
at the wider issues surrounding why so many nurses smoke. The importance of the large numbers who smoke before entering nursing has also been overlooked. Similarly, a number of studies have been conducted with the general smoking population, in an attempt to identify why adolescents and particularly women smoke. Although a number of studies reported weakly that adolescents felt that smoking relieved stress, few have explored the young person's perception of stress. Moreover there are only a limited number of studies which have explored to any great extent the relationship between the smoking behaviour of young women or indeed women in general and stress in the working environment (Daykin, 1993; Sheahan and Lattimer, 1995). Daykin (1993) surveyed 25 young women and 10 young men one year after entering the labour market as a follow up to a previous study which dealt with broad issues of health and work. Daykin reported that before leaving school, many of the young people had experimented with smoking. At the time of the follow up interviews 10 young women and three young men smoked regularly. The report concentrates on the young women's accounts of smoking which identi®ed links between smoking and both employment status and the experiences of family life. Overall young women reported that con¯icts with parents had increased since leaving school. These con¯icts surrounded issues such as the degree of autonomy allowed to young women and disputes over housework. Stress in the workplace was perceived as frustrations in relation to job insecurity, boredom and feelings of subordination. Although only a small study the ®ndings do provide important insights into the smoking behaviour of young adolescents and particularly young women. It is clear that on entering employment the young women in this study not only had to learn to cope with feelings of subordinance and lack of autonomy but also domestic and cultural constraints of femininity were in tension with the expectations of becoming an independent adult. Sheahan and Lattimer (1995) used 25 min telephone interviews to survey the smoking behaviour of 332 women aged 18±55 years in the USA. The aim of the study was to determine the prevalence of smoking and the relationships between stress, depression and emotional support and smoking. The ®ndings revealed that 28% (n = 92) of those surveyed were smokers. Of this ®gure, 63% were employed and 27% solely supported their families. The researchers reported that while there was no signi®cant dierence between smoker and nonsmoker in relation to emotional support, smokers had signi®cantly higher stress and depression scores. Forty-seven per cent reported that work stress was a very important contributing factor to their
smoking behaviour and another 21% perceived it to be a moderately important factor. Moreover, 29% stated that home stress was a very important factor in¯uencing their smoking, while a further 32% felt it was a moderately important factor. While the ®ndings suggested a link between stress in the workplace and smoking, it is clear that home stress also contributed considerably to the smoking behaviour of women in this study. Similar to many of the nursing studies, the report does not identify at what age the women took up smoking and whether they smoked prior to commencing employment. Studies on adolescents and females in the general smoking population provide a useful backdrop against which to compare the smoking behaviour of nurses. The ®ndings share a number of similarities and suggest that nurses are subject to the same kinds of stresses as other females and adolescents and thus they smoke for the same reasons. A few studies have alluded to the fact that the smoking behaviour of nurses is in¯uenced by peers who smoke. This work is reviewed alongside studies carried out among young adolescents in general, and females in particular who smoke.
3. The in¯uence of peer and social in¯uences on smoking amongst nurses Wagner (1985) surveyed the smoking behaviour of 504 registered nurses in New York. Using a questionnaire approach, a 62% response rate was achieved. Twenty-eight per cent of the sample self reported that they were smokers. Corresponding with most of the above studies, the majority of nurses who smoked did so prior to commencing nursing with 43% taking up the habit during their training. Seventy-one per cent reported that the main reason they commenced smoking was because their friends smoked thus highlighting the power of peer/social in¯uences on smoking behaviour. Similarly, Blakey and Seaton (1992) surveyed the smoking behaviour of 649 student nurses and pupil nurses in order to determine the reasons why nurses smoke. They achieved a 95% response rate and reported that 33.1% of the sample were smokers. Moreover, 75% of those who smoked commenced smoking before the age of 18 years. Relatively few had given up during their training. Eighty-two per cent of the students stated that they tended to smoke more when with a friend, suggesting that a peer/social in¯uence is relevant to the smoking behaviour of student nurses. Elkind (1988) also found that the majority of student nurses who smoked had smoking friends. The work of both Blakey and Seaton (1992) and Elkind
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
(1988) supports the association of friendship groups and nurses' smoking. Rowe (1998) surveyed the smoking behaviour of 555 quali®ed nurses in one hospital and 263 student nurses in one college of nursing in Northern Ireland and found that 21% of the quali®ed nurses and 46% of the students were smokers. The majority of quali®ed nurses (62%) commenced their smoking habit before coming to nursing with 89% of them being in¯uenced by their peers as adolescents. Moreover, almost all (95.5%) of students commenced smoking prior to commencing nursing. As with the quali®ed nurses, the major in¯uence which initiated smoking was that of peer pressure (77%). To explore further the power of peer relationships in the initiation and maintenance of smoking amongst nurses, Rowe conducted audio recorded semi structured interviews with 53 student nurses and 29 quali®ed nurses. She found that the overriding in¯uence on student nurses in continuing to smoke was that of smoking peers. The power of the relationship with smoking friends was so strong that eorts to stop smoking were seen as contrary to the views of other members within the group. Students found it easier to continue to smoke than to face the hostility shown to them by their smoking friends when trying to quit. This is consistent with ®ndings from studies of adolescents' smoking behaviour in general. For example, Van Roosmalen and McDaniel (1992) rigorously investigated the smoking behaviour of 1689 adolescents in the general population in Canada. Their ®ndings also indicated that the peer environment was a crucial factor in the smoking behaviour of both female and male adolescents. To remain an acceptable member of the peer group, it was essential to conform to the smoking practices of that group. Similarly, Fergusson et al. (1995) surveyed the smoking behaviour of 1256 young people. The ®ndings indicated that those who showed an early predisposition to cigarette smoking, also showed clear tendencies to aliate with like-minded peers and in turn, peer aliations during adolescence were one of the strongest and most consistent predictors of cigarette smoking. Both males (56.6%) and females (62.4%) smoked more often when they had a best friend who smoked regularly. It would appear that the stronger the friendship bond with smoking peers, the less likely it is that young people will stop smoking. These studies oer important insights into the power of the relationship of peer and social in¯uence on the smoking behaviour of adolescents in general and females in particular. Exploring the smoking behavior of adolescents in Great Britain, Diamond and Goddard (1995) report the ®ndings of a large scale survey carried out by the Department of Health (1994). The study examined the smoking behaviour among secondary school children in England, Scotland and Wales. Similar to the above
177
studies, the ®ndings identify that in all three countries there was a clear association between pupils' smoking behaviour and that of their friends. In England 75% of regular smokers said all or most of their friends smoked. In Wales the percentage was similar (71%) and likewise, in Scotland (73%). The studies support the peer ``bonding'' that exists among adolescents who smoke. In general, studies which have examined the reasons why adolescents and students smoke provide a useful backdrop to the research which examines student nurses' smoking behaviour. Previous studies have identi®ed that a large majority of student nurses commence smoking prior to coming to nursing and in the main few give up as they progress through training (Booth and Faukner, 1986; Carmichael and Cockcroft, 1990). It would therefore seem likely that the smoking behaviour of student nurses is con®rmed and maintained through friendship networks which formulate within nursing. Thus it would appear that the reasons why student nurses smoke, are similar to those of other adolescents and students in the general population who smoke.
4. The in¯uence of educationl attainment and type of course on smoking behaviour A few general adolescent smoking studies have looked at educational attainment at school as a predictor of smoking. The general idea is that smokers tend to be ``fed up'' with school, are underachievers and often leave school by the age of sixteen (Charlton, 1984). For example, Krohn et al. (1986) surveyed the smoking behaviour of 1800 grade 9±12 students in Iowa. Using a questionnaire approach a response rate of 91% was achieved. The aim was to to examine if adolescent social disaection with school was a signi®cant predictor of cigarette use. They found that smoking was higher among students who were less successful and less interested in school and school related activities. Krohn et al. also found that 73.1% of those who were most poorly integrated into the school smoked at least once a month, whilst only 27.6% of those well integrated into the school, smoked at all. While the ®ndings indicated that low achievement was associated with smoking, Krohn et al. reported that the most important factor in the smoking behaviour of adolescents was that of friendship groups. It would seem that this may indeed be the starting point in understanding the smoking behaviour of adolescents and may override other in¯uences. Charlton and Blair (1989) explored the smoking habits of 2885 young people aged 12 and 13 years in England. A questionnaire was administered to all,
178
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
under examination conditions supervised by their class teachers. This was followed 4 months later by completion of a second identical questionnaire. The aim of the study was to investigate whether absence from school could be predicted on the basis of their smoking behaviour. The ®ndings revealed that 20% of the young people were absent from school during the time of the second questionnaire. Of those absent, 47% of boys were regular smokers and 40% of girls were regular smokers. The most common reasons given for absenteeism were minor health ailments, such as common colds, headaches and stomach upsets. Charlton and Blair were unable to determine with any relevant degree of accuracy whether absenteeism was due to genuine health problems or to a generally negative attitude to school in the group who were smoking. They suggest that young people who smoke and are absent from school with minor ailments might be underachievers because of their smoking, rather than the fact that they take up smoking because they are underachievers. Thus the idea that a large majority of smokers are deviants who see school as an unnecessary preparation for the jobs they aspire to when they leave school (Eckert, 1983) is not supported by this study. Similarly, Nutbeam and Aaro (1991) surveyed the smoking behaviour of school children aged 11±16 years. The study drew upon data from the WHO cross national study undertaken in 1985±1986 in 11 countries. Data was elicited through a self-completed questionnaire administered in the classroom. The response rate varied between 87.8% and 96.6%. The aim of the study was to explore the relationship between negative attitudes towards school and regular cigarette smoking. Nutbeam and Aaro reported that a greater number of smokers was found among those who considered their achievement to be only average or below. In spite of varying prevalences of smoking across the countries, the higher the degree of school alienation, the higher degree of weekly smokers. Nutbeam and Aaro however, do emphasise the need for more empirical work in this area in order to understand more clearly the reasons why smokers reject school. A number of studies have also examined the level of education of the student nurse and `type' of training to determine if in fact it does have any in¯uence on the number of nurses that smoke. Harris et al. (1987) surveyed the smoking behaviour of 204 student nurses in two models of nurse training in Australia gaining a response rate of 93%. They report that while fewer nurses smoked in the hospital based training (31%), compared to the college based student nurses (38%), there was no statistically signi®cant dierence between the extent of smoking amongst student nurses within the two groups. Consistent with previous ®ndings from nurses' smoking studies, 75% of the student nurses began smoking before they commenced training. Har-
ris et al. conclude that the factors which lead most student nurses to smoke, lie in their experiences prior to nursing and not in the nature or location of their training programme. Similarly Rausch et al. (1987) examined the smoking behaviour of 555 senior student nurses enrolled in diploma, associate degree and degree programmes, within 11 nursing schools in the state of Alabama. The overall prevalence of smoking among the nurses in the sample was 26% but these ®ndings should be treated with caution given the very low response rate of 32% and the failing to provide a de®nition of `smoker'. Once more, the ®ndings indicate that there was no statistically signi®cant dierence in smoking prevalence between levels of nursing education. In contrast to the previous two studies, Elkind (1988) compared the smoking behaviour of a group of student teachers and a group of student nurses. Elkind claims that dierences in education contribute to occupational variations in the smoking behaviour of both groups of students. The ®ndings demonstrate that more student nurses (33%) were smoking at entry to their training than were student teachers (16%). Elkind however, acknowledges that this cannot be explained by the age or sex structures of the occupations as according to the OPCS (1977), both groups (teaching and nursing) were considered as being in the junior non-manual socio-economic group. The student nurses' smoking behaviour however, re¯ected that of women in general in this group, while the prevalence among the student teachers was more like that of women categorised as professional such as doctors and solicitors. It would seem that the association of lower educational attainment and higher prevalence of smoking, may have been made on the basis of the dierent educational attainment required for entry to both courses. For example, in the main, students are required to attain `A' level grades to pursue a career in teaching, while student nurses require a lower level of attainment. Elkind reported that while nurse training did not lead to greater initiation of smoking than teacher training, entry to nurse training was more likely than entry to teacher training, to lead to con®rmation and consolidation of the smoking habit with few nurses giving up the habit. This further demonstrates that there is a need for studies of nurses' smoking to be viewed within the context of the general adolescent smoking population. While authors have oered a variety of suggestions about why nurses smoke, these ideas are not substantiated. There are con¯icting explanations involving stress in the workplace, peer and social in¯uence and lower educational attainment. Given that the majority of nurses are female, it would seem important to consider the smoking behaviour of nurses in the context of gender (Graham, 1993).
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
5. Smoking and gender issues Many of the studies reviewed above, point to explanations for smoking which lie outside the occupational context but may relate to other factors including gender. Graham (1989) stresses the need for qualitative studies which explore the place of smoking in the lives of women, especially women with dual roles. This is supported by Adriaanse et al. (1991) who suggest that research into nurses' smoking should be based on comprehensive designs, taking into account life-style variables, as well as variables and characteristics associated with the work environment. In comparing the dierences between women and men's smoking behaviour, researchers have suggested that women are more motivated to smoke for pleasure and reduction of negative aect (Livson and Leino, 1988; Graham, 1993). Livson and Leino (1988) obtained data from 109 men and women, in an attempt to determine gender and psychological dierences pertaining to motive for smoking. Six commonly found motivating factors for smoking were identi®ed as reduction of negative aect, habit, addiction, pleasure, stimulation and sensorimotor manipulation. Signi®cant gender dierences were reported for two of the motives, that of reduction of negative aect and pleasure and this suggests that more women than men smoke for these reasons. They also speculate that there has been a societal modi®cation in women's roles during the past few decades, with more women taking up employment. Despite this change, they have had minimal relief from the `traditional' burden of maintaining the home and rearing children and thus, they smoke for relief of stress and pleasure. Graham's (1993) in-depth qualitative study develops this theory. She explored the link between cigarette smoking and dimensions of mothers' everyday lives. The study which was longitudinal in design employed interviews to obtain data from 905 mothers. The mothers were interviewed in their homes when their baby was six months old. Graham found that cigarettes were associated not only with the maintenance of normal caring routines but were also part of the way women coped with breakdown in these patterns of caring. She suggests that smoking is deeply embedded in the past and present lives of many mothers in working class households. Graham acknowledged the fact that self-reported data may potentially re¯ect subjective assessments of caring responsibilities, material circumstances and smoking status. In general however, the ®ndings illuminate the association of smoking with women's normal routine of living and provide important insight into the role that cigarettes have, in women with caring responsibilities. Issues of gender have also been identi®ed as being of
179
considerable importance in explaining the disparity of career opportunities aorded to female nurses compared to the male population of nurses (Porter, 1992). Porter implies that this practice has put considerable stress on the female sector of the profession but there is a lack of empirical evidence. While epidemiological research points to the importance of understanding the links between women, social class, educational status and smoking (Jacobson, 1986; Graham, 1989), few studies have looked in-depth into the relationship between women's smoking, their employment status and their domestic responsibilities. The lack of systematically collected data in studies of nurses' smoking has also hindered analysis of nurses' smoking behaviour, and reinforces the necessity of measuring all variables, including those speci®cally related to gender. In an attempt to understand why quali®ed nurses continue to smoke, in spite of being oered a supportive smoking cessation programme, Rowe (1998) conducted 29 indepth semi-structured audio recorded interviews with quali®ed nurses and found that their smoking behaviour mirrored that of mothers surveyed in Graham's (1993) study. Smoking was a vital part of the nurses' everyday lives and was inextricably linked to stress related to a multiplicity of roles. In order to cope with stress which was linked to being the main bread winner, coupled with having domestic, family and professional responsibilities, cigarettes had become the nurses' best friend and dependable lifeline. Quali®ed nurses gave vivid accounts of their love and deep passion for cigarettes. It is clear from Rowe's study that nurses' smoking behaviour must be explored within the context of their everyday lived experiences and not within the context of the nursing environment alone. 6. Summary In summary, this review of the literature demonstrates that there is limited up-to-date information which oers explanations of nurses' smoking behaviour. Given the importance attached to the nurse's health professional role and the pressure for nurses and other health professionals to act as good role models, there is a need to gain a clearer understanding of the factors in¯uencing their smoking behaviour. The literature reviewed in this article points to the dangers of linking smoking amongst nurses to the occupation of nursing. Rowe's (1998) study indicates, that the experiences, perceptions and behaviour of quali®ed nurses and student nurses who smoke mirror those of women and young people in general and they should be explored and understood in this context. The fact that many have initiated this unhealthy lifestyle behaviour before coming to nursing has not been systematically addressed nor have the reasons for
180
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
maintaining smoking been adequately explored. This de®cit is compounded by the fact that the majority of research focusing on nurses has not kept pace with studies which have concentrated on the general smoking population as a whole. The reason why many nurses smoke cannot therefore be determined with any degree of accuracy. While there is some evidence to suggest that nurses, student nurses in particular, feel that the nursing environment is potentially stressful, there has not been a clear link between stress in the workplace and smoking. These ®ndings correspond with those from general female adolescent studies. Indeed, it would seem that smoking amongst young people is initated and maintained through peer friendship groups which network in response to frustrations and feelings of subordination. It is likely then, that the reason nurses smoke, is no dierent to that of female adolescents in general who smoke. Moreover, studies which have attempted to show whether nurses' smoking is in¯uenced by level of education or type of course have not clearly demonstrated this link. Nurses fall into socio-economic group II (Oce for National Statistics, 1996) and in the light of this, there is a need to examine the smoking behaviour of quali®ed nurses in a demographic context. The majority of nurses are female and therefore it is necessary to consider their smoking behaviour alongside that of women in the general population. Furthermore, the majority of student nurses are in the 18±25 year age group and the majority of those who smoke choose to do so before coming to nursing. This being the case, it is essential to examine student nurses' smoking behaviour alongside that of adolescents and young adults in general. 6.1. Summary of review 1. Stress amongst nurses: nurses are subject to the same kinds of stresses as other females and adolescents and thus smoke for the same reasons. 2. Peer and social in¯uences: a large majority of nurses commence smoking prior to coming to nursing. 3. In¯uence of educational attainment and type of training: there are con¯icting explanations with regard to type of training and nurses smoking. 4. Gender issues: nurses smoking must be explored within the context of their everyday lived experiences and not within the nursing environment alone.
References Adriaanse, H., Reek, J.V., Sandbelt, L., Evers, G., 1991. Nurses smoking worldwide. A review of 73 surveys on
nurses' tobacco consumption in 21 countries in the period 1959±1988. International Journal of Nursing Studies 28 (4), 361±375. Becker, D.M., Myers, A.H., Sacci, M., Weida, S., Swank, R., Levine, D.M., Pearson, T.A., 1986. Smoking behaviour and attitudes toward smoking among hospital nurses. American Journal Public Health 76 (12), 1449±1451. Blakey, R., Seaton, A., 1992. Attitude amongst nursing tutors and their students health. Health Bulletin 50 (6), 417±421. Booth, K., Faulkner, A., 1986. Links between nurses and cigarette smoking. Nurse Education Today 6, 176± 182. Carmichael, A., Cockcroft, A., 1990. Survey of student nurses' smoking habits in a London teaching hospital. Respiratory Medicine 84, 277±282. Cinelli, B., Glover, E.D., 1988. Nurses smoking in the workplace: causes and solutions. Journal of Community Health Nursing 5 (4), 255±261. Charlton, A., 1984. The Brigantia Smoking Survey: a general review. Public Education About Cancer 77, 92±102. Charlton, A., Blair, V., 1989. Absence from school related to children's and parental smoking habits. British Medical Journal 298, 90±92 14 January. Charlton, A., While, D., Mochizuki, Y., 1997. A survey into the smoking habits of nursing students. Nursing Times 93 (39) 24 September. Daykin, N., 1993. Young women and smoking: towards a sociological account. Health Promotion International 8 (2), 95±102. Department of Health and Social Services, 1994. Regional Trends. Central Statistical Oce, HMSO, London. Diamond, A., Goddard, E., 1995. Smoking Among Secondary School Children in 1994. An enquiry carried out by Social Survey Division of OPCS on behalf of the Department of Health, The Welsh Oce and the Scottish Home and Health Department, London, HMSO. Eckert, P., 1983. Beyond the statistics of adolescent smoking. American Journal of Public Health 73 (4), 439±441 April. Elkind, A.K., 1988. Do nurses smoke because of stress? . Journal of Advanced Nursing 13, 733±745. Faulkner, A., Ward, L., 1983. Nurses as health educators in relation to smoking. Nursing Times 79 (15), 47±48. Feldman, B.B., Richard, E., 1986. Prevalence of nurse smokers and variables identi®ed with successful and unsuccessful smoking cessation. Res. Nurs. Health 9, 131±138. Fergusson, D.M., Lynskey, M.T., Horwood, L.J., 1995. The role of peer aliations, social, family and individual factors in continuities in cigarette smoking between childhood and adolescence. Addiction 90, 647±659. Graham, H., 1989. Women and smoking in the United Kingdom: the implications for health promotion. Health Promotion 3 (4), 371±382. Graham, H., 1993. When Life's a Drag: Women, Smoking and Disadvantage. HMSO Department of Health, London. Harris, M., Woodward, A., Bond, M., 1987. Do tertiary trained nurses smoke less than hospital trained nurses? . Community Health Studies XI (Suppl.), 41±44. Jacobson, B., 1986. Beating the Ladykillers: Women and Smoking. Pluto, London. Krohn, M.D., Naughton, M.J., Skinner, W.F., Becker, S.L.,
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181 Lauer, R.M., 1986. Social disaection, friendship patterns and adolescent cigarette use: the Muscatine Study. Journal of School Health 56 (4), 146±150. Leathar, D.S., 1980. Smoking among student nurses. Nursing Times 76 (14), 589±590. Livson, N., Leino, E.V., 1988. Cigarette smoking motives: factorial structure and gender dierences in a longitudinal study. The International Journal of Addiction 23 (6), 535± 544. Macleod Clark, J., Haverty, S., Kendall, S., 1987. Helping Patients and Clients to Stop Smoking, Health Education Authority Research Report No. 19 London, Health Education Authority. Nutbeam, D., Aaro, L.E., 1991. Smoking and pupil attitudes towards school: the implications for health education with young people. Results from the WHO study of health behaviour among school children. Health Education Research 6 (4), 415±421. Oce for Population Consensus and Surveys, 1977. Smoking and Professional People. HMSO, London. Oce for National Statistics, 1996. Living in Britain Results from 1996 General Household Survey Standard Occupational Classi®cation, Volume 3 HMSO. Porter, S., 1992. Women in a women's job: the gendered ex-
181
perience of nurses. Sociology of Health and Illness 13 (4), 510±527. Rausch, J.C., Zimmerman, G., Hopp, J., 1987. Smoking behaviour of student nurses enrolled in diploma, associate degree and undergraduate nursing programmes. Journal of Advanced Nursing 12, 111±119. Rowe, K., 1998. Nurses Who Smoke: The Experience of Attempting to Quit. Unpublished thesis submitted for the degree of PhD King's College University of London. Sheahan, S.L., Lattimer, M., 1995. Correlates of smoking stress and depression among women. Health Values 19 (1), 29±36 January/February. Spencer, J.K., 1984. Nurses, cigarette smoking in England and Wales. International Journal of Nursing Studies 21 (2), 69±79. Tagliacozzo, R., Vaughn, S., 1982. Stress and smoking in hospital nurses. American Journal Public Health 5, 441± 448. Van Roosmalen, E.H., McDaniel, S.A., 1992. Adolescent smoking intentions: gender dierences in peer context. Adolescence 27 (105), 87±105. Wagner, T.G., 1985. Smoking behaviour of nurses in western New York. Nursing Research 34 (1), 58±60 January/ February.
www.elsevier.com/locate/ijnurstu
Why nurses smoke: a review of the literature Kathy Rowe a,*, Jill Macleod Clark b a
The School of Nursing and Midwifery, The Queen's University of Belfast, University Road, Belfast, Northern Ireland, UK b University of Southampton, Southampton, UK Received 2 March 1999; accepted 6 June 1999
Abstract The smoking behaviour of nurses has been widely debated in the context of their professional role and responsibilities. There has been much speculation about why nurses smoke and possible explanations include a stressful nursing environment, peer pressure and socio economic status and education. This paper provides an overview of the literature which oers insights into the reasons why nurses smoke and compares the ®ndings from this literature with those studies examining the smoking behaviour of women in general and young women in particular. This review reveals that many students take up smoking before commencing their training and the factors which in¯uence nurses smoking are similar to those that in¯uence similar groups of females in the general population. # 2000 Published by Elsevier Science Ltd. All rights reserved. Keywords: Nurses smoking; Women smoking; Student smoking; Adolescent smoking
1. Introduction Over the years there has been an ongoing debate about the implications of the smoking behaviour of nurses in relation to their own health and their health promotion role. (Faulkner and Ward, 1983; Macleod Clark et al., 1987; Rowe, 1998). Empirical data about the smoking behaviour of nurses is limited with many studies being out of date or based on small samples (Adriaanse et al., 1991; Rowe, 1998). There seems to be a general consensus from the literature that the incidence of smoking amongst nurses equates with that of women in the population in general, and is around
* Corresponding author. Tel.: +44-1232-273997; fax: +441232-319356. E-mail address: [email protected] (K. Rowe).
28% (Oce for National Statistics, 1996). This rate in nursing is in¯uenced by status and professional role, with those in more senior managerial positions smoking less than student nurses, enrolled nurses, new sta nurses and nursing auxiliaries (Becker et al., 1986; Feldman and Richard, 1986). This again equates with variations in smoking rates in the general female population which relate to occupational status (Oce for National Statistics, 1996). A number of explanations have been oered in an attempt to establish why nurses smoke. The three main themes that emerge from a review of the literature are those of stress, peer and social in¯uence and socio economic status and education. The review which is presented under these three headings was undertaken using a number of computer data bases but with particular scrutiny of CINAHL 1982±1997 and Medline 1992±1997.
0020-7489/00/$ - see front matter # 2000 Published by Elsevier Science Ltd. All rights reserved. PII:S 0 0 2 0 - 7 4 8 9 ( 9 9 ) 0 0 0 6 0 - 7
174
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
2. The in¯uence of stress on smoking amongst nurses Many studies of nurses' smoking behaviour have attempted to demonstrate that smoking is a coping mechanism against stress, caused by the nursing environment (Leathar, 1980; Tagliacozzo and Vaughn, 1982; Spencer, 1984; Booth and Faulkner, 1986; Elkind, 1988; Cinelli and Glover, 1988; Carmichael and Cockcroft, 1990; Charlton et al., 1997). While a number of these studies support this theory, con¯icting ®ndings are also reported. The studies are presented in chronological order and re¯ect the smoking behaviour of student nurses1 pupil nurses2 and quali®ed nurses. Leathar (1980) surveyed the smoking behaviour of 93 nurses of all grades in order to investigate the reasons why they smoked. He employed in-depth interviews, each lasting at least 1 h and focused on understanding nurses' smoking in relation to stress speci®c to nursing. Participants were encouraged to identify particular areas of diculty, such as death, night duty, a rigid authority structure and responsibility problems. Leathar found a crucial time appeared to be during the ®rst 18 months of training when some nurses started smoking and others who already smoked increased their levels of smoking. A number of reasons were given for smoking, such as not knowing anyone and having to make friends, having to cope with more responsibility than one should for her status particularly on night duty and the general lack of support given by trained sta. In order to cope with the continual uncertainty, frustration and lack of support, informal friendship groups were formed and cigarettes were used as a symbol of friendship and mutual trust. While the ®ndings suggest that student nurses smoked in order to cope with stress incurred in the nursing environment, the report does not identify how many nurses were in fact smoking prior to commencing nursing. Moreover, the study does not probe to any extent the power of peer friendship groups in initiating and maintaining smoking. Tagliacozzo and Vaughn (1982) surveyed 933 registered nurses in seven hospital departments to establish whether there was an association between a stressful nursing environment and nurses' smoking behaviour. They reported that 19.9% of nurses smoked and that nurses who perceived their job as stressful were more likely to smoke than nurses who did not. It is not 1 Student nurse: applies to a student who is enrolled on a 3year programme for the professional quali®cation of registered nurse. 2 Pupil nurse: applies to a student who is enrolled on a 2year programme for the professional quali®cation of enrolled nurse. This programme is no longer available in the UK.
reported however, how many nurses were smoking prior to their entry to nursing. The ®ndings indicated that nurses most likely to perceive the physical and emotional demands of the job as stressful, were under 29 years and single. The authors acknowledge the need for further research to explore whether smoking amongst younger nurses is a mechanism for coping, or whether younger nurses who smoke are simply more likely than older nurses to perceive work situations as stressful. However, given the limited response rate (49.3%) and the use of an ambiguous de®nition of smoker the ®ndings should be interpreted with some caution. Spencer's (1984) survey of the smoking behaviour of 1577 nurses in two health authorities in England and Wales also attempted to determine whether smoking was related to stress in the nursing environment. The sample comprised all types of quali®ed nurses, student nurses and nursing assistants. The results revealed that although nurses did ®nd particular parts of their job stressful, most coped well with this stress. Therefore the idea that either the feelings of stress associated with particular nursing tasks, or the ability to cope with the stress, is connected to cigarette smoking was not supported by his evidence. In fact the ®ndings identify that more non-smokers (55.5%) felt stressed by patients dying, than did smokers (46.8%). In all other activities the amount of stress felt by smokers and non smokers was very similar, thus the view that nursing, cigarette smoking and stress are connected, may not be supported. Booth and Faulkner (1986) surveyed the smoking behaviour of 563 student and pupil nurses and found that 36.4% were smoking. Moreover 73% of the current smokers claimed to have been smoking for longer than three years. Thus the researchers concluded that since 95% of the sample were still in training, most were smoking prior to coming to nursing. In order to determine some of the factors associated with smoking, they moved beyond the questionnaire data and conducted in-depth interviews with 10% of the nurses who smoked. This data revealed that over half of the smokers in the sample perceived worry, stress or tension as contributing to their smoking behaviour. Despite this, it is interesting to note that very few (4.4%) of current smokers reported speci®c stressful nursing situations as being associated with increased smoking. In fact only nine nurses mentioned nursing at all and four of these appeared to interweave nursing with outside factors in their lives. These ®ndings are similar to those elicited by Elkind (1988) and Rowe (1998). They add weight to the view that while nurses use cigarettes as a coping strategy for the stressful situations they encounter in their everyday lives, their smoking behaviour should be considered in a wider context
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
than that of stress generated by the nursing environment alone. Elkind (1988) surveyed 43 undergraduate student teachers3 and 69 learner nurses4, during their ®rst year of training. The aim of the study was to explore the smoking behaviour of student nurses and teachers at entry to training, and to determine the factors which underlie the initiation of smoking. The design of the study included a baseline questionnaire administered at commencement of training, followed by an interview during the induction term. Three further questionnaires were distributed during the ®rst year of training at 5 months, 8±10 months and 12±15 months after entry to their course, followed by a second interview. The ®ndings revealed that 33% of student nurses were smokers at entry to their nursing programme, compared with 16% of student teachers. This again supports the view that a large number of nurses do commence their smoking habit prior to commencing their nursing career. Pre-eminent among the student nurses and student teachers responses (smokers and non-smokers alike), was the idea that smoking relieved stress (52%). The concept of stress was further explored during the second interview and it appeared that both nurses and student teachers alike experienced a signi®cantly greater degree of stress when in the practical environment than in the academic environment. While Elkind concludes that stress among student nurses is an issue, it would seem that patterns of smoking behaviour may be established at entry to training and therefore the role of stress should not be interpreted over simplistically. Cinelli and Glover (1988) surveyed the smoking behaviour of 299 registered nurses in Pennsylvania utilising an adaptation of the Job-related Stress and Smoking Behaviour Questionnaire developed by Tagliacozzo and Vaughn (1982). The questionnaire which was revised to include other sections such as career history, smoking history, personal data and nurse job stress index, yielded a response rate of 49%. The purpose of the study was to examine the level of professional nurse training of nurses who smoked, the aect of social factors on the initiation and reinforcement of smoking in the workplace and home, and the impact of occupational stress on smoking behaviour. The authors reported that occupational stress was not signi®cantly related to the smoking behaviour of nurses, and the strongest predictor for smoking amongst nurses was in fact the number of friends who 3 Student teacher: refers to a student enrolled on a programme for the professional quali®cation of degree in education. 4 Learner nurses: refers to students enrolled on both registered nurse and enrolled nurse programmes.
175
smoked. A current smoker was de®ned as an individual who smoked more than one tobacco cigarette a day for a period of three months or longer during the past year. This is an unusual de®nition and no explanation is given as to why individuals who smoked one cigarette a day were excluded from the study. Again these ®ndings should be treated with some caution given the poor response rate. Carmichael and Cockcroft (1990) used a questionnaire to examine the smoking behaviour of 368 student nurses and found that 43% of the sample were smokers. They further reported that 82% of those who were smoking were smoking on entry to nursing. This once more highlights the importance of exploring the reasons why nurses smoke within a much wider context than the nursing environment. While Carmichael and Cockcroft reported that stress, greater opportunity and peer group behaviour in¯uenced 18% of student nurses to commence smoking and in¯uenced a further 18% to increase the number they smoked, they did not explore the reasons why the vast majority (82%) were smoking before coming to nursing. A stressful nursing environment cannot therefore be implicated as a major reason why nurses smoke. Charlton et al. (1997) more recently explored the smoking behaviour of 90 student nurses at classes in one university. A questionnaire approach was used and a response rate of 72.2% was achieved. The ®ndings revealed that of those surveyed, 28% were current smokers. The researchers however, acknowledge that during distribution of the questionnaire, the absence rate from class was high due to a ¯u epidemic. It is therefore possible that this percentage may not re¯ect an accurate picture of the extent of student nurses smoking in this university. Although this is a small scale study, the ®ndings are similar to that of previous studies, in that the majority of smoking students were smoking prior to coming to nursing. In fact 64% had tried their ®rst cigarette before the age of 16 years and the rest between the ages of 16 and 18 years. Moreover, levels of stress were not signi®cantly associated with smoking behaviour. This once again, suggests that the assumption about stress and smoking amongst nurses needs to be revisited, particularly given the numbers of students who bring their smoking to nursing. In general, studies which have examined the reasons why nurses smoke do seem to identify that the nursing environment is potentially stressful but they fail to comprehensively demonstrate a clear link between stress in the workplace and smoking. Findings from some of the studies are compromised by methodological weaknesses including low response rates and lack of detail to de®nition of a smoker. These researchers have attempted to examine the relationship between stress in the workplace and smoking and failed to look
176
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
at the wider issues surrounding why so many nurses smoke. The importance of the large numbers who smoke before entering nursing has also been overlooked. Similarly, a number of studies have been conducted with the general smoking population, in an attempt to identify why adolescents and particularly women smoke. Although a number of studies reported weakly that adolescents felt that smoking relieved stress, few have explored the young person's perception of stress. Moreover there are only a limited number of studies which have explored to any great extent the relationship between the smoking behaviour of young women or indeed women in general and stress in the working environment (Daykin, 1993; Sheahan and Lattimer, 1995). Daykin (1993) surveyed 25 young women and 10 young men one year after entering the labour market as a follow up to a previous study which dealt with broad issues of health and work. Daykin reported that before leaving school, many of the young people had experimented with smoking. At the time of the follow up interviews 10 young women and three young men smoked regularly. The report concentrates on the young women's accounts of smoking which identi®ed links between smoking and both employment status and the experiences of family life. Overall young women reported that con¯icts with parents had increased since leaving school. These con¯icts surrounded issues such as the degree of autonomy allowed to young women and disputes over housework. Stress in the workplace was perceived as frustrations in relation to job insecurity, boredom and feelings of subordination. Although only a small study the ®ndings do provide important insights into the smoking behaviour of young adolescents and particularly young women. It is clear that on entering employment the young women in this study not only had to learn to cope with feelings of subordinance and lack of autonomy but also domestic and cultural constraints of femininity were in tension with the expectations of becoming an independent adult. Sheahan and Lattimer (1995) used 25 min telephone interviews to survey the smoking behaviour of 332 women aged 18±55 years in the USA. The aim of the study was to determine the prevalence of smoking and the relationships between stress, depression and emotional support and smoking. The ®ndings revealed that 28% (n = 92) of those surveyed were smokers. Of this ®gure, 63% were employed and 27% solely supported their families. The researchers reported that while there was no signi®cant dierence between smoker and nonsmoker in relation to emotional support, smokers had signi®cantly higher stress and depression scores. Forty-seven per cent reported that work stress was a very important contributing factor to their
smoking behaviour and another 21% perceived it to be a moderately important factor. Moreover, 29% stated that home stress was a very important factor in¯uencing their smoking, while a further 32% felt it was a moderately important factor. While the ®ndings suggested a link between stress in the workplace and smoking, it is clear that home stress also contributed considerably to the smoking behaviour of women in this study. Similar to many of the nursing studies, the report does not identify at what age the women took up smoking and whether they smoked prior to commencing employment. Studies on adolescents and females in the general smoking population provide a useful backdrop against which to compare the smoking behaviour of nurses. The ®ndings share a number of similarities and suggest that nurses are subject to the same kinds of stresses as other females and adolescents and thus they smoke for the same reasons. A few studies have alluded to the fact that the smoking behaviour of nurses is in¯uenced by peers who smoke. This work is reviewed alongside studies carried out among young adolescents in general, and females in particular who smoke.
3. The in¯uence of peer and social in¯uences on smoking amongst nurses Wagner (1985) surveyed the smoking behaviour of 504 registered nurses in New York. Using a questionnaire approach, a 62% response rate was achieved. Twenty-eight per cent of the sample self reported that they were smokers. Corresponding with most of the above studies, the majority of nurses who smoked did so prior to commencing nursing with 43% taking up the habit during their training. Seventy-one per cent reported that the main reason they commenced smoking was because their friends smoked thus highlighting the power of peer/social in¯uences on smoking behaviour. Similarly, Blakey and Seaton (1992) surveyed the smoking behaviour of 649 student nurses and pupil nurses in order to determine the reasons why nurses smoke. They achieved a 95% response rate and reported that 33.1% of the sample were smokers. Moreover, 75% of those who smoked commenced smoking before the age of 18 years. Relatively few had given up during their training. Eighty-two per cent of the students stated that they tended to smoke more when with a friend, suggesting that a peer/social in¯uence is relevant to the smoking behaviour of student nurses. Elkind (1988) also found that the majority of student nurses who smoked had smoking friends. The work of both Blakey and Seaton (1992) and Elkind
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
(1988) supports the association of friendship groups and nurses' smoking. Rowe (1998) surveyed the smoking behaviour of 555 quali®ed nurses in one hospital and 263 student nurses in one college of nursing in Northern Ireland and found that 21% of the quali®ed nurses and 46% of the students were smokers. The majority of quali®ed nurses (62%) commenced their smoking habit before coming to nursing with 89% of them being in¯uenced by their peers as adolescents. Moreover, almost all (95.5%) of students commenced smoking prior to commencing nursing. As with the quali®ed nurses, the major in¯uence which initiated smoking was that of peer pressure (77%). To explore further the power of peer relationships in the initiation and maintenance of smoking amongst nurses, Rowe conducted audio recorded semi structured interviews with 53 student nurses and 29 quali®ed nurses. She found that the overriding in¯uence on student nurses in continuing to smoke was that of smoking peers. The power of the relationship with smoking friends was so strong that eorts to stop smoking were seen as contrary to the views of other members within the group. Students found it easier to continue to smoke than to face the hostility shown to them by their smoking friends when trying to quit. This is consistent with ®ndings from studies of adolescents' smoking behaviour in general. For example, Van Roosmalen and McDaniel (1992) rigorously investigated the smoking behaviour of 1689 adolescents in the general population in Canada. Their ®ndings also indicated that the peer environment was a crucial factor in the smoking behaviour of both female and male adolescents. To remain an acceptable member of the peer group, it was essential to conform to the smoking practices of that group. Similarly, Fergusson et al. (1995) surveyed the smoking behaviour of 1256 young people. The ®ndings indicated that those who showed an early predisposition to cigarette smoking, also showed clear tendencies to aliate with like-minded peers and in turn, peer aliations during adolescence were one of the strongest and most consistent predictors of cigarette smoking. Both males (56.6%) and females (62.4%) smoked more often when they had a best friend who smoked regularly. It would appear that the stronger the friendship bond with smoking peers, the less likely it is that young people will stop smoking. These studies oer important insights into the power of the relationship of peer and social in¯uence on the smoking behaviour of adolescents in general and females in particular. Exploring the smoking behavior of adolescents in Great Britain, Diamond and Goddard (1995) report the ®ndings of a large scale survey carried out by the Department of Health (1994). The study examined the smoking behaviour among secondary school children in England, Scotland and Wales. Similar to the above
177
studies, the ®ndings identify that in all three countries there was a clear association between pupils' smoking behaviour and that of their friends. In England 75% of regular smokers said all or most of their friends smoked. In Wales the percentage was similar (71%) and likewise, in Scotland (73%). The studies support the peer ``bonding'' that exists among adolescents who smoke. In general, studies which have examined the reasons why adolescents and students smoke provide a useful backdrop to the research which examines student nurses' smoking behaviour. Previous studies have identi®ed that a large majority of student nurses commence smoking prior to coming to nursing and in the main few give up as they progress through training (Booth and Faukner, 1986; Carmichael and Cockcroft, 1990). It would therefore seem likely that the smoking behaviour of student nurses is con®rmed and maintained through friendship networks which formulate within nursing. Thus it would appear that the reasons why student nurses smoke, are similar to those of other adolescents and students in the general population who smoke.
4. The in¯uence of educationl attainment and type of course on smoking behaviour A few general adolescent smoking studies have looked at educational attainment at school as a predictor of smoking. The general idea is that smokers tend to be ``fed up'' with school, are underachievers and often leave school by the age of sixteen (Charlton, 1984). For example, Krohn et al. (1986) surveyed the smoking behaviour of 1800 grade 9±12 students in Iowa. Using a questionnaire approach a response rate of 91% was achieved. The aim was to to examine if adolescent social disaection with school was a signi®cant predictor of cigarette use. They found that smoking was higher among students who were less successful and less interested in school and school related activities. Krohn et al. also found that 73.1% of those who were most poorly integrated into the school smoked at least once a month, whilst only 27.6% of those well integrated into the school, smoked at all. While the ®ndings indicated that low achievement was associated with smoking, Krohn et al. reported that the most important factor in the smoking behaviour of adolescents was that of friendship groups. It would seem that this may indeed be the starting point in understanding the smoking behaviour of adolescents and may override other in¯uences. Charlton and Blair (1989) explored the smoking habits of 2885 young people aged 12 and 13 years in England. A questionnaire was administered to all,
178
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
under examination conditions supervised by their class teachers. This was followed 4 months later by completion of a second identical questionnaire. The aim of the study was to investigate whether absence from school could be predicted on the basis of their smoking behaviour. The ®ndings revealed that 20% of the young people were absent from school during the time of the second questionnaire. Of those absent, 47% of boys were regular smokers and 40% of girls were regular smokers. The most common reasons given for absenteeism were minor health ailments, such as common colds, headaches and stomach upsets. Charlton and Blair were unable to determine with any relevant degree of accuracy whether absenteeism was due to genuine health problems or to a generally negative attitude to school in the group who were smoking. They suggest that young people who smoke and are absent from school with minor ailments might be underachievers because of their smoking, rather than the fact that they take up smoking because they are underachievers. Thus the idea that a large majority of smokers are deviants who see school as an unnecessary preparation for the jobs they aspire to when they leave school (Eckert, 1983) is not supported by this study. Similarly, Nutbeam and Aaro (1991) surveyed the smoking behaviour of school children aged 11±16 years. The study drew upon data from the WHO cross national study undertaken in 1985±1986 in 11 countries. Data was elicited through a self-completed questionnaire administered in the classroom. The response rate varied between 87.8% and 96.6%. The aim of the study was to explore the relationship between negative attitudes towards school and regular cigarette smoking. Nutbeam and Aaro reported that a greater number of smokers was found among those who considered their achievement to be only average or below. In spite of varying prevalences of smoking across the countries, the higher the degree of school alienation, the higher degree of weekly smokers. Nutbeam and Aaro however, do emphasise the need for more empirical work in this area in order to understand more clearly the reasons why smokers reject school. A number of studies have also examined the level of education of the student nurse and `type' of training to determine if in fact it does have any in¯uence on the number of nurses that smoke. Harris et al. (1987) surveyed the smoking behaviour of 204 student nurses in two models of nurse training in Australia gaining a response rate of 93%. They report that while fewer nurses smoked in the hospital based training (31%), compared to the college based student nurses (38%), there was no statistically signi®cant dierence between the extent of smoking amongst student nurses within the two groups. Consistent with previous ®ndings from nurses' smoking studies, 75% of the student nurses began smoking before they commenced training. Har-
ris et al. conclude that the factors which lead most student nurses to smoke, lie in their experiences prior to nursing and not in the nature or location of their training programme. Similarly Rausch et al. (1987) examined the smoking behaviour of 555 senior student nurses enrolled in diploma, associate degree and degree programmes, within 11 nursing schools in the state of Alabama. The overall prevalence of smoking among the nurses in the sample was 26% but these ®ndings should be treated with caution given the very low response rate of 32% and the failing to provide a de®nition of `smoker'. Once more, the ®ndings indicate that there was no statistically signi®cant dierence in smoking prevalence between levels of nursing education. In contrast to the previous two studies, Elkind (1988) compared the smoking behaviour of a group of student teachers and a group of student nurses. Elkind claims that dierences in education contribute to occupational variations in the smoking behaviour of both groups of students. The ®ndings demonstrate that more student nurses (33%) were smoking at entry to their training than were student teachers (16%). Elkind however, acknowledges that this cannot be explained by the age or sex structures of the occupations as according to the OPCS (1977), both groups (teaching and nursing) were considered as being in the junior non-manual socio-economic group. The student nurses' smoking behaviour however, re¯ected that of women in general in this group, while the prevalence among the student teachers was more like that of women categorised as professional such as doctors and solicitors. It would seem that the association of lower educational attainment and higher prevalence of smoking, may have been made on the basis of the dierent educational attainment required for entry to both courses. For example, in the main, students are required to attain `A' level grades to pursue a career in teaching, while student nurses require a lower level of attainment. Elkind reported that while nurse training did not lead to greater initiation of smoking than teacher training, entry to nurse training was more likely than entry to teacher training, to lead to con®rmation and consolidation of the smoking habit with few nurses giving up the habit. This further demonstrates that there is a need for studies of nurses' smoking to be viewed within the context of the general adolescent smoking population. While authors have oered a variety of suggestions about why nurses smoke, these ideas are not substantiated. There are con¯icting explanations involving stress in the workplace, peer and social in¯uence and lower educational attainment. Given that the majority of nurses are female, it would seem important to consider the smoking behaviour of nurses in the context of gender (Graham, 1993).
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
5. Smoking and gender issues Many of the studies reviewed above, point to explanations for smoking which lie outside the occupational context but may relate to other factors including gender. Graham (1989) stresses the need for qualitative studies which explore the place of smoking in the lives of women, especially women with dual roles. This is supported by Adriaanse et al. (1991) who suggest that research into nurses' smoking should be based on comprehensive designs, taking into account life-style variables, as well as variables and characteristics associated with the work environment. In comparing the dierences between women and men's smoking behaviour, researchers have suggested that women are more motivated to smoke for pleasure and reduction of negative aect (Livson and Leino, 1988; Graham, 1993). Livson and Leino (1988) obtained data from 109 men and women, in an attempt to determine gender and psychological dierences pertaining to motive for smoking. Six commonly found motivating factors for smoking were identi®ed as reduction of negative aect, habit, addiction, pleasure, stimulation and sensorimotor manipulation. Signi®cant gender dierences were reported for two of the motives, that of reduction of negative aect and pleasure and this suggests that more women than men smoke for these reasons. They also speculate that there has been a societal modi®cation in women's roles during the past few decades, with more women taking up employment. Despite this change, they have had minimal relief from the `traditional' burden of maintaining the home and rearing children and thus, they smoke for relief of stress and pleasure. Graham's (1993) in-depth qualitative study develops this theory. She explored the link between cigarette smoking and dimensions of mothers' everyday lives. The study which was longitudinal in design employed interviews to obtain data from 905 mothers. The mothers were interviewed in their homes when their baby was six months old. Graham found that cigarettes were associated not only with the maintenance of normal caring routines but were also part of the way women coped with breakdown in these patterns of caring. She suggests that smoking is deeply embedded in the past and present lives of many mothers in working class households. Graham acknowledged the fact that self-reported data may potentially re¯ect subjective assessments of caring responsibilities, material circumstances and smoking status. In general however, the ®ndings illuminate the association of smoking with women's normal routine of living and provide important insight into the role that cigarettes have, in women with caring responsibilities. Issues of gender have also been identi®ed as being of
179
considerable importance in explaining the disparity of career opportunities aorded to female nurses compared to the male population of nurses (Porter, 1992). Porter implies that this practice has put considerable stress on the female sector of the profession but there is a lack of empirical evidence. While epidemiological research points to the importance of understanding the links between women, social class, educational status and smoking (Jacobson, 1986; Graham, 1989), few studies have looked in-depth into the relationship between women's smoking, their employment status and their domestic responsibilities. The lack of systematically collected data in studies of nurses' smoking has also hindered analysis of nurses' smoking behaviour, and reinforces the necessity of measuring all variables, including those speci®cally related to gender. In an attempt to understand why quali®ed nurses continue to smoke, in spite of being oered a supportive smoking cessation programme, Rowe (1998) conducted 29 indepth semi-structured audio recorded interviews with quali®ed nurses and found that their smoking behaviour mirrored that of mothers surveyed in Graham's (1993) study. Smoking was a vital part of the nurses' everyday lives and was inextricably linked to stress related to a multiplicity of roles. In order to cope with stress which was linked to being the main bread winner, coupled with having domestic, family and professional responsibilities, cigarettes had become the nurses' best friend and dependable lifeline. Quali®ed nurses gave vivid accounts of their love and deep passion for cigarettes. It is clear from Rowe's study that nurses' smoking behaviour must be explored within the context of their everyday lived experiences and not within the context of the nursing environment alone. 6. Summary In summary, this review of the literature demonstrates that there is limited up-to-date information which oers explanations of nurses' smoking behaviour. Given the importance attached to the nurse's health professional role and the pressure for nurses and other health professionals to act as good role models, there is a need to gain a clearer understanding of the factors in¯uencing their smoking behaviour. The literature reviewed in this article points to the dangers of linking smoking amongst nurses to the occupation of nursing. Rowe's (1998) study indicates, that the experiences, perceptions and behaviour of quali®ed nurses and student nurses who smoke mirror those of women and young people in general and they should be explored and understood in this context. The fact that many have initiated this unhealthy lifestyle behaviour before coming to nursing has not been systematically addressed nor have the reasons for
180
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181
maintaining smoking been adequately explored. This de®cit is compounded by the fact that the majority of research focusing on nurses has not kept pace with studies which have concentrated on the general smoking population as a whole. The reason why many nurses smoke cannot therefore be determined with any degree of accuracy. While there is some evidence to suggest that nurses, student nurses in particular, feel that the nursing environment is potentially stressful, there has not been a clear link between stress in the workplace and smoking. These ®ndings correspond with those from general female adolescent studies. Indeed, it would seem that smoking amongst young people is initated and maintained through peer friendship groups which network in response to frustrations and feelings of subordination. It is likely then, that the reason nurses smoke, is no dierent to that of female adolescents in general who smoke. Moreover, studies which have attempted to show whether nurses' smoking is in¯uenced by level of education or type of course have not clearly demonstrated this link. Nurses fall into socio-economic group II (Oce for National Statistics, 1996) and in the light of this, there is a need to examine the smoking behaviour of quali®ed nurses in a demographic context. The majority of nurses are female and therefore it is necessary to consider their smoking behaviour alongside that of women in the general population. Furthermore, the majority of student nurses are in the 18±25 year age group and the majority of those who smoke choose to do so before coming to nursing. This being the case, it is essential to examine student nurses' smoking behaviour alongside that of adolescents and young adults in general. 6.1. Summary of review 1. Stress amongst nurses: nurses are subject to the same kinds of stresses as other females and adolescents and thus smoke for the same reasons. 2. Peer and social in¯uences: a large majority of nurses commence smoking prior to coming to nursing. 3. In¯uence of educational attainment and type of training: there are con¯icting explanations with regard to type of training and nurses smoking. 4. Gender issues: nurses smoking must be explored within the context of their everyday lived experiences and not within the nursing environment alone.
References Adriaanse, H., Reek, J.V., Sandbelt, L., Evers, G., 1991. Nurses smoking worldwide. A review of 73 surveys on
nurses' tobacco consumption in 21 countries in the period 1959±1988. International Journal of Nursing Studies 28 (4), 361±375. Becker, D.M., Myers, A.H., Sacci, M., Weida, S., Swank, R., Levine, D.M., Pearson, T.A., 1986. Smoking behaviour and attitudes toward smoking among hospital nurses. American Journal Public Health 76 (12), 1449±1451. Blakey, R., Seaton, A., 1992. Attitude amongst nursing tutors and their students health. Health Bulletin 50 (6), 417±421. Booth, K., Faulkner, A., 1986. Links between nurses and cigarette smoking. Nurse Education Today 6, 176± 182. Carmichael, A., Cockcroft, A., 1990. Survey of student nurses' smoking habits in a London teaching hospital. Respiratory Medicine 84, 277±282. Cinelli, B., Glover, E.D., 1988. Nurses smoking in the workplace: causes and solutions. Journal of Community Health Nursing 5 (4), 255±261. Charlton, A., 1984. The Brigantia Smoking Survey: a general review. Public Education About Cancer 77, 92±102. Charlton, A., Blair, V., 1989. Absence from school related to children's and parental smoking habits. British Medical Journal 298, 90±92 14 January. Charlton, A., While, D., Mochizuki, Y., 1997. A survey into the smoking habits of nursing students. Nursing Times 93 (39) 24 September. Daykin, N., 1993. Young women and smoking: towards a sociological account. Health Promotion International 8 (2), 95±102. Department of Health and Social Services, 1994. Regional Trends. Central Statistical Oce, HMSO, London. Diamond, A., Goddard, E., 1995. Smoking Among Secondary School Children in 1994. An enquiry carried out by Social Survey Division of OPCS on behalf of the Department of Health, The Welsh Oce and the Scottish Home and Health Department, London, HMSO. Eckert, P., 1983. Beyond the statistics of adolescent smoking. American Journal of Public Health 73 (4), 439±441 April. Elkind, A.K., 1988. Do nurses smoke because of stress? . Journal of Advanced Nursing 13, 733±745. Faulkner, A., Ward, L., 1983. Nurses as health educators in relation to smoking. Nursing Times 79 (15), 47±48. Feldman, B.B., Richard, E., 1986. Prevalence of nurse smokers and variables identi®ed with successful and unsuccessful smoking cessation. Res. Nurs. Health 9, 131±138. Fergusson, D.M., Lynskey, M.T., Horwood, L.J., 1995. The role of peer aliations, social, family and individual factors in continuities in cigarette smoking between childhood and adolescence. Addiction 90, 647±659. Graham, H., 1989. Women and smoking in the United Kingdom: the implications for health promotion. Health Promotion 3 (4), 371±382. Graham, H., 1993. When Life's a Drag: Women, Smoking and Disadvantage. HMSO Department of Health, London. Harris, M., Woodward, A., Bond, M., 1987. Do tertiary trained nurses smoke less than hospital trained nurses? . Community Health Studies XI (Suppl.), 41±44. Jacobson, B., 1986. Beating the Ladykillers: Women and Smoking. Pluto, London. Krohn, M.D., Naughton, M.J., Skinner, W.F., Becker, S.L.,
K. Rowe, J. Macleod Clark / International Journal of Nursing Studies 37 (2000) 173±181 Lauer, R.M., 1986. Social disaection, friendship patterns and adolescent cigarette use: the Muscatine Study. Journal of School Health 56 (4), 146±150. Leathar, D.S., 1980. Smoking among student nurses. Nursing Times 76 (14), 589±590. Livson, N., Leino, E.V., 1988. Cigarette smoking motives: factorial structure and gender dierences in a longitudinal study. The International Journal of Addiction 23 (6), 535± 544. Macleod Clark, J., Haverty, S., Kendall, S., 1987. Helping Patients and Clients to Stop Smoking, Health Education Authority Research Report No. 19 London, Health Education Authority. Nutbeam, D., Aaro, L.E., 1991. Smoking and pupil attitudes towards school: the implications for health education with young people. Results from the WHO study of health behaviour among school children. Health Education Research 6 (4), 415±421. Oce for Population Consensus and Surveys, 1977. Smoking and Professional People. HMSO, London. Oce for National Statistics, 1996. Living in Britain Results from 1996 General Household Survey Standard Occupational Classi®cation, Volume 3 HMSO. Porter, S., 1992. Women in a women's job: the gendered ex-
181
perience of nurses. Sociology of Health and Illness 13 (4), 510±527. Rausch, J.C., Zimmerman, G., Hopp, J., 1987. Smoking behaviour of student nurses enrolled in diploma, associate degree and undergraduate nursing programmes. Journal of Advanced Nursing 12, 111±119. Rowe, K., 1998. Nurses Who Smoke: The Experience of Attempting to Quit. Unpublished thesis submitted for the degree of PhD King's College University of London. Sheahan, S.L., Lattimer, M., 1995. Correlates of smoking stress and depression among women. Health Values 19 (1), 29±36 January/February. Spencer, J.K., 1984. Nurses, cigarette smoking in England and Wales. International Journal of Nursing Studies 21 (2), 69±79. Tagliacozzo, R., Vaughn, S., 1982. Stress and smoking in hospital nurses. American Journal Public Health 5, 441± 448. Van Roosmalen, E.H., McDaniel, S.A., 1992. Adolescent smoking intentions: gender dierences in peer context. Adolescence 27 (105), 87±105. Wagner, T.G., 1985. Smoking behaviour of nurses in western New York. Nursing Research 34 (1), 58±60 January/ February.