... Employment and Sickness Absence. ... The same factors influence job turnover and long spells of sick leave a 3-year follow-up of Swedish nurses

Transcription

1 European Journal of Public Health, Vol. 18, No. 4, ß The Author Published by Oxford University Press on behalf of the European Public Health Association. All rights reserved. doi: /eurpub/ckn009 Advance Access published on 21 February Employment and Sickness Absence... The same factors influence job turnover and long spells of sick leave a 3-year follow-up of Swedish nurses Malin Josephson 1, Per Lindberg 2, Margaretha Voss 2, Lars Alfredsson 3,4, Eva Vingård 1 Background: In many countries, a general shortage of nurses is a public health problem, and retention of nurses in active work is a challenge. The aim of this study was to ascertain whether the same individual factors, working conditions and health problems had led to increased probability of both leaving jobs and prolonged sickness absence in a cohort of Swedish nurses over a period of 3 years. Methods: A baseline questionnaire was answered by 2293 nurses, representing a response rate of 86%. Exposed and unexposed nurses were compared with regard to two outcomes. During the 3-year follow-up, exposed and unexposed nurses were compared with regard to two outcomes: resigning and having at least one sick leave spell that lasted 28 days or longer. Results: We found that 18% of the nurses left their employment, and 16% had sick leave spells 28 days. Work in geriatric care, being socially excluded by superiors and/or workmates, negative effects of organizational changes and poor self-rated general health were factors that increased the likelihood of both leaving jobs and long-term sick leave. Conclusions: The present results underline the importance of improving working conditions and supporting sustainable health in order to prevent high turnover and prolonged sick leave among nurses. Resigning and moving to another institution can be interpreted as a way to actively cope with an unhealthy work environment. Keywords: job turnover, nurses, resignations, sick leave, working conditions... Introduction n many countries, a general shortage of nurses is a public Ihealth problem, and retaining those health care professionals in active work and in the care organization constitutes a challenge for all stakeholders in society. 1 5 Resigning is one way for nurses to avoid negative conditions in their work environment. An earlier study of nurses, who have actually left their positions in health care showed that lack of professional opportunities, restricted professional autonomy and unsatisfactory salary contributed to the decision to quit. 6 Factors that were identified as predictors of leaving nursing and changing to other occupations included being young, being male, working as an assistant nurse, having musculoskeletal problems and having limited access to transfer devices for lifting patients. 7 Furthermore, it has been established that poor job satisfaction contributes to a general intention to leave the workplace or the profession in question Another way for nurses to avoid both the task load and the detrimental conditions in their workplaces is to be on prolonged sickness leave. Musculoskeletal disorders and mental problems are the most frequent diagnoses for longterm sickness absence in Sweden, as in many other countries. 12 Physical demands at work are still common, and nursing often 1 Occupational and Environmental Medicine, Uppsala University, Uppsala, Sweden 2 Section of Personal Injury Prevention, Karolinska Institutet, Stockholm, Sweden 3 Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden 4 Stockholm Center of Public Health, Karolinska University Hospital, Stockholm, Sweden Correspondence: Malin Josephson, Occupational and Environmental Medicine, Uppsala University, Akademiska sjukhuset, Uppsala, Sweden, tel: , Malin.Josephson@medsci.uu.se implies heavy lifting and awkward working positions, which are risk factors for low back pain. 13 A longitudinal study of Norwegian nurses showed that frequent handling of heavy objects predicts disabling low back pain and sick leave, and lack of encouragement and support in the workplace represents a risk factor for sickness absence. 14 Other studies of nurses have revealed that sick leave is associated with job strain, 15 diminished support from workmates and less time to plan work, 16 as well as a low level of perceived fairness in decision-making and negative social relations 17 and bullying 18 at the workplace. Nursing is often a stressful occupation that involves substantial demands and in many cases also insufficient resources, which are two risk factors for exhaustion and burnout. 19 According to the investigations cited earlier, many closely related factors are associated with both leaving the job and long periods of sickness absence. However, those studies analysed different populations of nurses working under disparate conditions. Another investigation of employees in various professions has indicated that resignations and sick leave are not connected with the same factors. 20 If that is indeed the case, there is also a need for different strategies to support the retention of nurses. Thus, our aim was to ascertain whether the same individual factors, working conditions and health problems, had led to increased probability of both leaving jobs and prolonged sickness absence in a cohort of Swedish nurses over a period of 3 years. Methods This investigation was conducted within the framework of the Swedish longitudinal study entitled Work and Sustainable Health in the Public Sector in Sweden (designated HAKuL) 21 and in cooperation with a European research project called Nurses Early Exit (NEXT). 1 The overall aim of the HAKuL study was to strengthen sustainable health among public

2 Nurses job turnover and sick leave 381 employees and implement early rehabilitation for those with impaired work ability. The NEXT project examined the reasons for premature withdrawal from the nursing profession in 10 European countries. 1 In Sweden, registered nurses have a college education (BSc), and assistant nurses have upper secondary school training. Both types of nurses, aged years, were included in the present study, and they were recruited at different facilities: hospitals and psychiatric care units in three different county councils, all primary care centres in one county council and geriatric care units and home care units in five municipalities. The participating institutions represented different geographical areas, and they varied in size and hence also in organizational structure, ranging from small worksites with two or three district nurses to hospitals with several hundred employees. The nurses were permanently employed and were not on long-term sick leave (>90 days) at the time the information was collected. A baseline questionnaire survey was carried out in a stepwise manner from November 1999 to January The questionnaire was sent to a total of 2664 actively working registered nurses and assistant nurses, and it was answered by 2293 of the recipients, representing a response rate of 86%. Self-rated health The questionnaire included the SF 36 instrument concerning general health and pain. 22 The scores for the scale were calculated as proposed for SF 36 and ranged from 0 to 100 (indicating lowest to highest well-being). Poor self-reported health was classified as belonging approximately to the quartile with the lowest score, which in our sample was a score of 60 on the general health scale and 41 for pain. Job factors Perceived physical exertion was used as a proxy for the general physical workload. The employees were asked the following question: How physically demanding do you in general perceive your work to be? The response scale was numbered 6 20, and the values were accompanied by verbal clarifications ranging from resting to maximal exertion. 23 Those who reported physical exertion greater than hard (score 14) were categorized as exposed. Five items were used to estimate the mental workload and six items the decision latitude. 24 The four response options for each item ranged from No, almost never to Yes, often, and the score was divided approximately into quartiles. The highest quartile for demands (16 high demands) and the lowest quartile for decision latitude (15 low decision latitude) were regarded as exposed. Two items in the questionnaire concerned the interaction between work and private life: one asked whether the demands of work interfered with home and family life, and the other enquired whether home/family life had a negative impact on work. Responses were marked on a 5-point scale, ranging from very seldom (1) to very often (5). 25 Responding very often to one of the items was considered to represent a negative interaction between working and private life a work-family conflict. Consequences of organizational changes were assessed by the question How have you been influenced by the changes that have been made in your workplace over the past year? That query included four sub-items in the form of two positive statements ( It has been developing and instructive and It has entailed new opportunities for me ) and two negative statements ( I cannot carry out my work tasks as well as I want to and I am beginning to doubt whether I can manage ), to which the participants could respond that they agreed, partly agreed or did not agree. 26 If they agreed with at least one of the negative statements and with neither of the positive statements, they were categorized as reporting a negative development in their working conditions. Two questions concerned social exclusion at the workplace. Social exclusion was defined as a feeling that superiors and/or workmates to some degree shut their subordinates out by not greeting them, not answering their questions and not giving them information. Satisfaction with the quality of care was measured by the single question Are you generally satisfied with the care you can give patients at your workplace? Responses of never, seldom or sometimes were categorized as unsatisfactory quality of patient care, and often was considered to represent satisfactory care quality. The nurses were also asked whether it was possible for them to get a new job without moving. Recorded data In , data on sick leave spells 28 days and resignations were collected and recorded via continuous contact with the studied care facilities and their supervisors. Statistics The proportion of nurses who resigned or had long spells of sick leave are presented with 95% CI by sex, age, occupation and area of health care. Odds ratio (OR) values with 95% CI were calculated separately for termination of employment and prolonged sick leave by means of multivariate logistic regression analysis including all four determinants in the same model. The same analyses were also done including one work-related factor, all the work-related factors that were considered or one self-reported health factor at a time and in all cases adjusting for age and sex. Due to the intercorrelation between the variables, and also because of the possibility that some of the measured variables could be intervening factors in the causality between working conditions, health problems, sick leave and turnover, we did not perform the multivariate analyses using all explored variables in a single model. Results Over the 3-year study period, 18% of the investigated nurses resigned, and 16% had sick leave spells that lasted at least 28 days. Only 2% had both long spells of sick leave and left their jobs. Individual factors and area of health care A higher rate of leavers was observed among nurses younger than 50 years of age compared with those who were older, and also among registered nurses compared with assistant nurses. Considering long spells of sick leave, the pattern was the opposite: the rate of long spells was higher for older nurses, and also among assistant nurses compared with registered nurses. In addition, female nurses had a higher proportion of long sick leave spells than male nurses. Compared with other health care areas, nurses in geriatric and primary care had a higher rate of resignations, and they also tended to have a higher rate of long periods of sick leave(table 1). The ORs obtained in the multivariate analyses demonstrate the probability of nurses leaving the job or having long spells of sick leave, respectively. Younger nurses (50 years) and registered nurses showed an increased probability of resignations and a decreased likelihood of long sick spells. Moreover, the probability of long-term sickness absence was lower for the male nurses than for the females (OR 0.39, 95% CI ). Also, the differences between health care areas appeared mainly in the multivariate analyses, indicating increased probability of both resigning and long sick spells for those working in

3 382 European Journal of Public Health Table 1 Proportions of nurses who resigned from their jobs or had one or more sick leave spells lasting 28 days over a period of 3 years, analysed according to sex, age, type of nursing certificate and area of health care for employment Number Percentage (95% CI) who resigned Percentage (95% CI) with sick leave 28 days Sex Female ( ) 17.1 ( ) Male ( ) 7.4 ( ) Age <50 years ( ) 13.5 ( ) 50 years ( ) 20.7 ( ) Nursing certificate Assistant nurses ( ) 17.9 ( ) Registered nurses ( ) 14.2 ( ) Health care area Hospital care ( ) 14.0 ( ) Psychiatric care ( ,1) 14.5 ( ) Primary care ( ) 17.1 ( ) Geriatrics ( ) 18.4 ( ) Total probability of long-term sickness absence in the analyses adjusted for sex and age. The associations were diluted when all work related variables were considered together. More than a third of the nurses (35%) reported that they were only sometimes, seldom, or never satisfied with the quality of care they provided. High demands, dissatisfaction with the work schedule and work-family conflicts increased the odds of resigning, and the same tendency, with the exception of dissatisfaction with the schedule, was apparent regarding the probability of long-term sickness absence. In addition, strenuous physical exertion at work tended to raise the odds of resigning and increased the probability of long-term sickness absence (table 3). Possibilities of getting a new job Being able to get a new job without moving increased the odds of resigning (OR 2.17, 95% CI ) and tended to decrease the probability of long-term sick leave (OR 0.86, 95% CI ) (table 3). Twenty-eight percent of the registered nurses reported that it was possible for them to find new employment without moving, compared with 17% of the assistant nurses. Table 2 OR with 95% test-based CI calculated for leaving employment and for having long spells of sick leave by performing multivariate logistic regression analysis of four different factors: sex, age, type of nursing certificate and area of health care Determinant Leaving employment Long sick spells Sex Female 1 1 Male 1.19 ( ) 0.39 ( ) Age 50 years 1 1 <50 years 1.77 ( ) 0.60 ( ) Nursing certificate Assistant nurse 1 1 Registered nurse 1.72 ( ) 0.76 ( ) Health care area Hospital 1 1 Psychiatric facility 0.89 ( ) 1.04 ( ) Primary care 1.66 ( ) 1.27 ( ) Geriatrics 1.77 ( ) 1.39 ( ) geriatric care and primary care, although the estimate for sick leave in primary care was not statistically significant (table 2). Working conditions Table 3 shows results obtained using logistic regression models including one work-related factor at the time and one model including all work-related factors. It seems that adverse working conditions influenced the probability of both resigning and long spells of sick leave, as indicated by the observation that social exclusion in the workplace and negative consequences of changes increased the likelihood of both. These two factors were reported by 15 and 20% of the nurses, respectively. Of the nurses exposed to social exclusion, 26% resigned and 20% had at least one long sick leave spell. Of those who reported a negative development in their working conditions due to reorganizations, 23% resigned and 24% had at least one long sick leave spell. Dissatisfaction with patient care increased the odds of resigning, and the same tendency was apparent regarding the Self-reported health problems Self-reported health complaints influenced the likelihood of both resigning and long periods of sick leave. Considering individuals with pain, there was a marked increase in the probability of long-term leave but only a tendency towards increased odds of quitting the job (table 4). There was also an indication that both those with poor health and those belonging to the quartile with very good health resigned to a greater extent than all others. The proportion of nurses who resigned was 19% among those with very good health, 21% of those with poor health and 17% of other categories. Discussion We found that work in geriatric care, social exclusion in the workplace, negative consequences of organizational changes and poor self-rated general health were mutual factors that increased both the likelihood of resigning and the odds of being on long-term sick leave in the cohort of nurses included in our study. Considering the first of the mentioned factors, our results underline the challenge involved in retaining nurses in geriatric health care. In Sweden, there are problems in recruiting nurses to geriatrics, and the status of that professional arena is relatively low compared to other fields of care. At the same time, the elderly population is growing, and along with that the need for geriatric care. Among the nurses we studied, a feeling that their superiors and/or workmates to some degree excluded them by not greeting them, not answering their questions and not giving them information increased the probability of leaving the job and of having long spells of sickness absence. The concept of exclusion is a part of a workplace bullying situation, which the Swedish National Board of Occupational Safety and Health (Ordinance 1993:17) describes as employees being placed outside the workplace community. 27 In a previous report from the HAKuL study of female employees in various occupations in the public sector, workplace bullying was identified as a risk factor for long-term sick leave. 21 Other investigations have shown that such harassment imposes stress reactions, reduces mental well-being and leads to the need for sick leave. 18,28,29 Bullying is a common problem among nurses and seems to be linked to occupational culture. 30,31 To be able to retain nurses in care facilities, signs of bullying must be dealt

4 Nurses job turnover and sick leave 383 Table 3 OR with 95% test-based CI calculated for leaving employment and for long spells of sick leave by performing multivariate logistic regression analyses adjusted for age and sex, considering one work-related factor at a time or considering all work related variables Determinant No. exposed Leaving employment Long sick spells a b a b Demands at work Low to moderate High ( ) 1.06 ( ) 1.28 ( ) 0.84 ( ) Decision latitude Moderate to high Low ( ) 1.14 ( ) 1.18( ) 1.14 ( ) Satisfied with care provided Yes No ( ) 1.14 ( ) 1.24 ( ) 1.04 ( ) Social exclusion No Yes ( ) 1.70 ( ) 1.47 ( ) 1.36 ( ) Consequences of work changes Not negative Negative ( ) 1.29 ( ) 1.91 ( ) 1.67 ( ) Satisfied with work schedule Yes No ( ) 1.15 ( ) 1.08 ( ) 0.94 ( ) Physical exertion Not excessive Excessive ( ) 1.30( ) 1.57 ( ) 1.48 ( ) Work family conflict None or occasionaly Very often ( ) 1.16 ( ) 1.46 ( ) 0.92 ( ) Possibility of getting new job Limited Yes ( ) 2.17 ( ) 0.85 ( ) 0.86 ( ) a: Adjusted for age and sex, considering one work-related factor at a time b: Adjusted for age, sex, considering all work-related factors Table 4 OR with 95% test-based CI calculated for leaving employment and for having long spells of sick leave by performing multivariate logistic regression analysis of one self-reported health variable at a time and adjusting for age and sex Determinant Number exposed Leaving employment Long sick spells General health Good 1 1 Poor ( ) 2.16 ( ) Pain No 1 1 Yes ( ) 2.02 ( ) with and workplace mores that create such behaviour must be prevented. Negative consequences of organizational changes at work were markedly associated with long-term sick leave and with resigning. These findings agree with previous studies showing that deterioration of the psychosocial working conditions, such as reductions in job control and social support, imply a risk of health problems and increased sickness absence. 16,32,33 Poor general health and pain were chiefly associated with sick leave. Musculoskeletal complaints and general health problems are often related to reduced work ability and can make it more difficult to get a new job. Before nurses actually resign, it is almost certain that they consider their possibilities of getting new and better positions by leaving. 34 Turnover imposes costs on the organization, puts pressure on remaining nurses and makes it difficult to maintain the knowledge base and the quality of the care given. 35,36 On the other hand, a factor that often contributes to nurses leaving their jobs is the desire for professional development, 10,37 and a certain level of circulation of nurses among organizations and institutions is actually a sign of progress and flexibility. 38 To determine whether the nurses who resigned actually left their profession or if they moved to another institution, we collected additional data within the framework of the European NEXT study. 1 Our analysis showed that a majority of those who quit, i.e. 86% of the registered nurses and 68% of the assistant nurses in our study remained in nursing, 39 and those rates agree with previous reports concerning registered nurses. Official statistics in Sweden 40 show that registered nurses are not fleeing their profession, and consequently the turnover that is observed is primarily a question of moving between institutions. A meta-analysis of antecedent predictors of turnover has demonstrated that educated women show the same pattern as men, namely, they change jobs but stay in the labour market. 6 Turnover between institutions does not indicate that registered nurses are less active in the nursing profession, but it can be interpreted as an active coping behaviour leading away from an unhealthy psychosocial environment. We found that assistant nurses, but not registered nurses, had a higher rate of long-term sickness absence than of leaving the job. Furthermore, compared with the registered nurses, the assistant nurses reported that their possibilities of finding new employment without moving were more limited. It is plausible that the higher turnover rate among registered nurses than among assistant nurses gives the former group the opportunity to avoid bullying and adverse psychosocial conditions in the workplace. Being able to leave such negative work surroundings may prevent prolonged periods of sick leave. Being employed in a non-preferred workplace or profession

5 384 European Journal of Public Health seems to be associated with little opportunity for development, weak support from superiors and health problems. 41 In Sweden, assistant nurses constitute one of the occupational groups with the highest rates of extended sickness absence, and, together with home-based personal care workers, they represent the largest number of cases of long-term sick leave. 42 Methodological considerations The selection of organizations for participation in the HAKuLproject was not a random process. A long-standing commitment to the study was necessary, because it was necessary for the organizations to help with the following aspects: administration of questionnaires, involvement in the rehabilitation measures, reporting of employees on long-term sick leave and nurses who quit their jobs, receiving feedback from the surveys and hopefully also initiation of interventions. However, we do not think that this led to selection of workplaces that either functioned relatively well or had problems. When an organization decided to participate, all nursing workplaces within the organization were involved, for example, all elderly care units in the local authority or all psychiatric care facilities in the county council. Insufficient contrast in the exposure conditions is often a weakness in epidemiological studies and may lead to underestimation of the association between exposure and outcome. All workplaces included in the present study were in the public sector and comprised only two closely related occupations: registered nurses and assistant nurses. The distribution of work conditions was limited and the disparities in demands, control and physical requirements may were largely due to individual differences in work capacity and rating behaviour. Moreover, a differential misclassification, may have occurred in the form of over-reporting of adverse work conditions among nurses who intented to quit their jobs or had health problems, which would had resulted in overestimation of the association between determinants and outcome. In addition, it is possible that the health selection of nurses represented a source of underestimation of the influence of poor health on a decision to remain in or end employment. Nurses who were on longterm sick leave at the beginning of the study period were excluded from the baseline measurement, which gave a low number of disability pensioners over the period. Furthermore, our study included a few potential determinants of long-term sick leave or resigning, and it is obvious that other variables that were not covered in our investigation also have had an impact on those aspects in our cohort. Nurses with repeated shorter spells of sick leave or spells that were only somewhat shorter than 28 days were included in the comparison group. This may have diluted the contrast between those with and without long periods of sickness absence. On the other hand, using a cut-off point of 28 days probably increased the validity of the collected data by ensuring that must of the cases were actually reported. It should also be mentioned that, for several years, it has been compulsory for employers in Sweden to take actions and consider a plan for return to work for any employee who is on sick leave for a period of 4 weeks or longer. Another potential limitation of our study is that we did not have access to data on nurses who were receiving sickness benefit after resigning from their jobs. Conclusions Working in geriatric care, being subjected to social exclusion at the workplace, negative consequences of organizational changes and self-reported health problems were found to be predictors of both leaving the job and long-term sickness absence. The results underline the importance of improving working conditions and supporting sustainable health among nurses in order to prevent excessively high turnover and prolonged sick leave in this occupational group. Resigning can be interpreted as a way to actively cope with an unhealthy work environment. Our findings emphasize that, at least in Sweden, sickness absence constitutes a significant disruption in the supply of nurses. Acknowledgement The authors thank AFA Labour Market Insurance for funding the HAKuL Study. Conflicts of interest: None declared. Key points The supply of nurses is influenced by both job turnover and long-term sick leave. The likelihood of both resignation and prolonged sickness absence among nurses is increased by work in geriatric care, a negative social and organizational environment in the workplace and poor health. The results underline the importance of improving working conditions and promoting sustainable health in order to prevent a shortage of nurses. Leaving the job can be interpreted as a strategy by which nurses actively cope with detrimental working conditions. References 1 Hasselhorn H-M, Tackenberg P, Müller BH. Working conditions and intent to leave the profession among nursing staff in Europe. Stockholm: National Institute for Working Life, Janiszewski Goodin H. The nursing shortage in the United States of America: an integrative review of the literature. J Adv Nurs 2003;43: Mullen C. Commentary: an English perspective comments to the article The nursing shortage in the United States of America: an integrative review of the literature. J Adv Nurs 2003;43: Raeve PD. Commentary: an European perspective. Comments to the article The nursing shortage in the United States of America: an integrative review of the literature. J Adv Nurs 2003;43: Wickett D, McCutcheon H, Long L. Commentary: an Australien perspective comments to the article The nursing shortage in the United States of America: an integrative review of the literature. J Adv Nurs 2003;43: Fochsen G, Sjögren K, Josephson M, Lagerström M. Factors contributing to the decision to leave nursing care: a study among Swedish nursing personnel. J Nurs Manag 2005;13(4): Fochsen G, Josephson M, Hagberg M, et al. Predictors of leaving nursing care: a longitudinal study among Swedish nursing personnel. Occup Environ Med 2006;63(3): Griffeth R, Hom PW, Gaertner S. A meta-analysis of antecedents and correlates of employee turnover: update, moderator tests, and research implications for the next millenium. J Manage 2000;26: Hom PW, Kinicki A. Toward a greater understanding of how dissatisfaction drives employee turnover. Acad Manage J 2001;44: Lum L, Kervin J, Clark K, et al. Explaining nursing turnover intent: job satisfaction, pay satisfaction, or organizational commitment? J Organ Behav 1998;19: Morrell K. Towards a typology of nursing turnover: the role of shocks in nurses decisions to leave. J Adv Nurs 2004;49: The National Social Insurance Board. Socialförsäkringsboken 2004 (The book of social insurance 2004). Stockholm: Riksförsäkringsverket (The National Social Insurance Board), 2004.

6 Nurses job turnover and sick leave Lagerström M, Hansson TH, Hagberg M. Work-related low-back problems in nursing. Scand J Work Environ Health 1998;24: Eriksen W, Bruusgaard D, Knardahl S. Work factors as predictors of intense or disabling low back pain; a prospective study of nurses aides. Occup Environ Med 2004;61: Bourbonnais R, Mondor M. Job strain and sickness absence among nurses in the province of Québec. Am J Ind Med 2001;39: Petterson I-L, Hertting A, Hagberg L, Theorell T. Are trends in work and health conditions interrelated? A study of Swedish hospital employees in the 1990s. J Occup Health Psychol 2005;10(2): Elovainio M, Kivimäki M, Vahtera J. Organizational justice: evidence of a new psychosocial predictor of health. Am J Public Health 2002;92(1): Kivimäki M, Elovainio M, Vahtera J. Workplace bullying and sickness absence in hospital staff. Occup Environ Med 2000;57: Demerouti E, Bakker A, Nachreiner F, Schaufeli W. A model of burnout and life satisfaction amongs nurses. J Adv Nurs 2000;32(2): Lund T, Iversen L, Poulsen KB. Work environment factors, health, lifestyle and marital status as predictors of job change and early retirement in physically heavy occupations. Am J Ind Med 2001;40(2): Vingård E, Lindberg P, Josephson M, et al. Long-term sick-listing among women in the public sector and its associations with age, social situation, lifestyle and work factors. A three year follow up study. Scand J Public Health 2005;33(5): Sullivan M, Karlsson J, Ware J. The Swedish SF-36 health survey. Evaluation of data quality, scaling assumptions, reliability and construct validity across general populations in Sweden. Social Sci Med 1995;41: Borg G. Borg s perceived exertion and pain scales. USA: Human Kinetics, Karasek RA, Theorell T. Healthy work. New York: Basic Books, Dallner M, Elo A-L, Gamberale F, et al. Validation of the general nordic questionnaire (QPS Nordic) for psychological and social factors at work. Copenhagen: Nordic council of ministers, Report No: 2000: Härenstam A, Karlqvist L, Bodin L, et al. Patterns of working and living conditions: a holistic, multivaraiate approach to occupational health studies. Work and Stress 2003;17: The Swedish National Board of Occupational Safety and Health. Victimization at work. Ordinance AFS 1993:17. Stockholm: Swedish Work Environment Authority, Hansen AM, Hogh A, Persson R, et al. Bullying at work, health outcomes, and physiological stress response. J Psychosom Res 2006;60: Varita MA-L. Consequences of workplace bullying with respect to the well-being of its targets and the observers of bullying. Scand J Work Environ Health 2001;27(1): Lewis AM. Nurses bullying: organizational considerations in the maintenance and perpetration of health care bullying cultures. J Nurs Manag 2006;14: Randle J. Bullying in the nursing profession. J Adv Nurs 2003;43(4): Kivimäki M, Vahtera J, Pentti J, Ferrie JE. Factors underlying the effect of organisational downsizing on health of employees: longitudinal cohort study. Br Med J 2000;320: Vahtera J, Kivimäki M, Pentti J, Theorell T. Effect of change in the psychosocial work environment on sickness absence: a seven year follow up of initially healthy employees. J Epidemiol Community Health 2000;54(7): Maertz CP, Griffeth RW. Eight motivational forces and voluntary turnover: a theoretical synthesis with implications for research. J Manag 2004;30(5): Anthony M, Standing T, Glick J, et al. Leadership and nurse retention. J Nurs Adm 2005;35: Jones CB. The costs of nurse turnover. J Nurs Adm 2004;34: Dalton DR. Employee transfer and employee turnover: a theoretical and practical disconnect? J Organ Behav 1997;18: Sjögren K, Fochsen G, Josephson M, Lagerström M. Reasons for leaving nursing care and improvements needed for considering a return: a study among Swedish nursing personnel. Int J Nurs Studies 2005;42(7). 39 Josephson M, Fochsen G, Lindberg P, Vingård E. Work related factors contributing to the decision to leave the workplace: Results for Swedish registered nurses and assistant nurses. In: Hasselhorn H-M, Müller BH, Tackenberg P, editors. NEXT Scientific Report June University of Wuppertal, 2005, p National Board of Health and Welfare. NPS 2003: Tilläggsrapport Tillgång på sjuksköterskor, barnmorskor och läkare (Supply of registered nurses, midwives and physicians). Stockholm: Socialstyrelsen (National Board of Health and Welfare), Aronsson G, Göransson S. Permanent employment but not in a preferred occupation: psychological and medical aspects, research implications. J Occup Health Psychol 1999;4(2): AFA Labour Market Insurance, Sweden. Available from: ring.se (accessed on 4 June 2007). Received 2 July 2007, accepted 23 January 2008