Workplace violence is recognized as a serious issue among the nursing profession,1,2 linked to a range of negative consequences for the employee3,4 and reduced quality in patient care.5 Consequently, it is important to understand the factors that might lead to workplace violence among nurses to identify effective intervention and prevention approaches. Previous research has tended to focus on workplace bullying, rarely exploring workplace violence.6-9 Furthermore, in this research, various nursing groups are often examined together, despite differences in work conditions. It is rare that particular types of nurses and their exposure to violence are examined.10 Nurses caring for the elderly (aged care nurses) in particular are under substantial long-term demand pressures and appear to be at high risk of workplace violence.11-13 With an aging patient population,14 the need for nurses to care for these vulnerable patients continues to increase and is exacerbated by the complex and chronic nature of patients and patient-nurse ratios.14 Workplace bullying research outside nursing has highlighted that characteristics of the Job Demands-Resources (JD-R) model, including increased demands such as these, may lead to a greater likelihood of exposure.15-17 Furthermore, individual characteristics such as negative affectivity (NA) have been associated with workplace bullying among a range of professionals.18,19 Therefore, the overall aim of this study was to examine the utility of the JD-R model and NA as potential antecedents of workplace violence among nurses who care for the elderly. The focus of the study was on the victim of violence (the nurse) and their subjective perceptions of exposure.
The nature of workplace violence is primarily physically oriented, inclusive of perceived and/or actual physical harm.20 Violent behavior can include both single and/or repeated acts of behavior(s) such as physical assault, threat of assault, emotional abuse, and verbal sexual harrassment.20 Sources of workplace violence may be internal to the organization (ie, coworkers and/or supervisors) and/or external (ie, patients or their family and friends).20 Despite the increased risk of nurses who care for the elderly being exposed to such types of workplace violence, little is known about the antecedent factors of violence among this group of nurses.
Drawing on workplace bullying research, the JD-R model is often applied. This model proposes that differing levels of job demands and resources are associated with certain employee outcomes.21 Job demands are defined as the physical, social, and organizational aspects of a job that require sustained energy. Resources can include job control and social support. Job control is the level of autonomy an employee has over his/her job demands, whereas social aspects can include the support of supervisors, coworkers, or family and friends. According to the model, high-strain jobs are those characterized by high demands and low resources, leading to physical and psychological costs (eg, stress, burnout, exhaustion).21,22 In line with the JD-R model, increased job demands and decreased control and support resources have been linked to workplace bullying.15-17 It is thought that the features of high-strain jobs may lead to frustration, conflict, and deteriorating relationships that may escalate into bullying.23 Furthermore, high-strain jobs may lead to violations of existing social, organizational, and work-related norms (eg, making errors at work that lead to negative perceptions), which increase the likelihood of being a target of bullying.24 It may be worthwhile exploring whether similar social-situational factors are also relevant for workplace violence. Little research has used the JD-R model in understanding the types of workplace violence or examining at-risk nursing groups, such as those who care for older patients specifically.
Furthermore, individual differences in NA have been shown to be important in workplace bullying research18,25 and therefore may also influence perceptions of other negative interpersonal encounters such as workplace violence. Negative affectivity can be defined as an individual’s level of pervasive negative emotionality and self-concept, whereby those with high levels of NA are more likely to experience negative emotions toward themselves and the world.26,27 A potential explanation of the role of NA relationship in workplace bullying is that NA acts as a perceptual bias and forms part of a vicious cycle.18,19 That is, an individual with high levels of NA might experience more interpersonal conflict, leading to heightened distress, interpretations of the conflict as more negative than it is, and, consequently, greater negative emotions. Thus, NA may play similar roles in reports of various types of workplace violence for nursing, particularly contexts of caring for the elderly; however, this is yet to be explored.
Purpose and Hypotheses
The overall aim of this study was to examine the utility of the JD-R model and NA as potential social-situational and individual antecedent factors of various types of workplace violence within the context of nurses who care for the elderly. The following hypotheses were proposed:
1. The components of the JD-R model will be related to reports of workplace violence, whereby high job demands and low resources (job control and social support) will be linked to reports of various types of workplace violence.
2. Higher levels of NA will be linked to reports of workplace violence types.
Design and Sample
This was a cross-sectional survey study undertaken with nurses working in elderly care facilities across a medium to large Australian healthcare organization. The nature of care in these facilities was primary acute and long-term care. Nurses in these care settings were sent survey packs to their work addresses. Those consenting to participate did so by completing the survey and using a reply paid envelope to return the survey. Ethics approval was obtained from the healthcare organization and universities involved.
Two hundred sixty-nine aged care nurses returned their survey, reflecting a response rate of 57.5%. The sample consisted of mostly females (92.6%; n = 249), 40 years or older (81.8%; n = 220). The majority had worked for the healthcare organization for 9 years or less (74.3%; n = 200), mostly in a part-time role (71.7%; n = 233), on morning shifts (34.6%; n = 93). The study sample was representative of nurses who care for the elderly among the Australian nursing population based on similarities of key demographic characteristics.14
Workplace violence was measured using an adapted version of a scale developed by Hesketh et al.20 The scale required respondents to rate the frequency of violence they perceived themselves as experiencing in their past 5 work shifts across violence types (ie, physical assault, threat of assault, emotional abuse, and verbal sexual harassment) and sources (ie, patient, visitor/family member of a patient, coworker, and supervisor) using a 4-point scale (“never,” “1 time,” “2 times,” and “3 or more times”). A definition of violence and examples of each violence type were provided. Groupings across violence types were created by collapsing coworker and supervisor violence as sources internal to the organization and violence from the patient and patient’s visitor/family member as sources that were external.
Job demands were measured using an 11-item scale developed by Caplan et al28 rating perceptions of physical and psychological job demands on a 5-point scale (“very often/a great deal,” “fairly often/a lot,” “sometimes/some,” “occasionally/a little,” and “rarely/hardly any”). The Cronbach’s α for job demands was .90. A 9-item job control scale created by Karasek29 was adopted with responses rated on a 5-point scale (“strongly disagree,” “disagree,” “neither,” “agree,” “strongly agree”). The Cronbach’s α for job control was .77. Social support was assessed using a 4-item measure by Caplan et al28 rated on a 5-point scale (“very much,” “somewhat,” “a little,” “not at all,” and “don’t have any such person”). Each item required 3 responses in relation to social support levels from the immediate supervisor, coworkers, and family and friends, and responses were collapsed based on these sources to form 3 subscales. The Cronbach’s α’s for these scales were .89, .81, and .79, respectively.
Negative affectivity was measured using the NA subscale from the Positive and Negative Affect Schedule developed by Watson et al.27 Respondents rated the degree to which they had experienced 10 negative emotions in the past week using a 5-point scale (“very slightly or not at all,” “a little,” “moderately,” “quite a bit,” “very much”). The Cronbach’s α for NA was .91.
Data were analyzed with SPSS Statistics version 17.0 (Chicago, Illinois). Separate ordinal regressions were conducted to examine the antecedents of each of the violence categories. Violence responses were collapsed and coded for the ordinal regression analyses to ensure adequate sample sizes. Scores of “never” and “1 time” remained as they were; however, scores of “2 times” and “3 times or more” were collapsed into a “yes, frequently” group. Harman’s ex-post 1-factor test was conducted, as outlined in Podsakoff and Organ,30 to check for common method variance. The items for the variables’ scales were entered into an unrotated factor analysis, which revealed 5 factors, indicating that common method variance was not influential.
Rates of Workplace Violence
Table 1 outlines the frequency of reported workplace violence for the nurses in the sample. Whereas the majority reported no experiences, 36.4% reported external physical assault, 35.7% reported external threats of assault, and 28.6% reported external emotional abuse. Low rates of exposure were found for internal threat of assault (2.6%), internal physical assault (3.3%), and internal verbal sexual harassment (0.8%).
The results of the ordinal regression analyses are provided in Table 2. In terms of internal emotional abuse, NA was linked to high levels of this type of violence (χ21 [n = 210] = 17.90, P = .000). High job demands (χ21 [n = 212] = 16.55, P = .000), low job control (χ21 [n = 212] = 4.60, P = .032), and high NA (χ21 [n = 212] = 22.20, P = .000) were antecedents of external emotional abuse. High levels of job demands (χ21 [n = 211] = 15.85, P = .000) and high NA (χ21 [n = 211] = 4.21, P = .040) were antecedents for external threat of assault. High levels of job demands (χ21 [n = 212] = 18.97, P = .000), high outside work support (χ21 [n = 212] = 5.57, P = .018), and low job control (χ21 [n = 212] = 6.90, P = .009) were linked with external physical assault. Finally, high job demands (χ21 [n = 211] = 5.77, P = .009) and high outside work support (χ21 [n = 211] = 4.23, P = .040) were linked with external verbal sexual harassment. Regression analyses could not be conducted for internal threat of assault, internal physical assault, and internal verbal sexual harassment due to the low number of reported cases.
Post hoc estimates of the power analyses for the ordinal regressions, including checks bracketing the power of ordinal coding by using logistic or Poisson regression assumptions, indicated that the analyses presented here typically had a power of more than 0.95 (using G*Power 3).31 Furthermore, sensitivity analyses of the power indicated that under severe (hypothetical) conditions of constraint, such as having large portions of the target variable determined by a control variable and the worst rates of prevalence obtained in this study, the lowest power levels obtained were 0.86, suggesting that the analyses had appropriate levels of power.
This study investigated the utility of the JD-R model and NA as potential social-situational and individual antecedent factors of various types of workplace violence within the context of nurses who care for the elderly. Overall, the nurses in this study reported concerning levels of external emotional abuse (12.6%; n = 34), external threat of assault (13.4%; n = 36), and external physical assault (15.2%; n = 41), with approximately 13% to 15% indicating frequent occurrences of each of these forms of violence. That is, this study’s results confirm previous findings that violence is at high levels for nurses1,2 and specifically for nurses who care for the elderly,11-13 particularly considering the zero tolerance policies of violence in these healthcare environments.
In terms of the situational variables, job demands were linked to all of the externally sourced types of violence. This suggests a mechanism whereby heavily loaded nurses are under such demands that external parties may feel that they, or their family member or friend, have not received the care or service expected. These findings highlight the increased risks associated with features of high-strain jobs, whereby norms may be violated15,24 and/or negative feelings and interpersonal experiences may occur as a result,23 potentially leading to increased exposure to these violence acts. Job control was also significantly associated with external emotional abuse and external physical assault. The direction of these relationships indicates that when nurses have discretion over how to carry out their work, they experienced fewer violent incidents. Overall, these findings extend previous research exploring the JD-R and bullying15-17 to aspects of workplace violence in an aged care nursing context.
The more person-based resources in the situation, reflecting social support, were only significant for outside work support onto external physical assault and external sexual harassment. The direction of these relationships was negative. These findings suggest either the causality is reversed, where these forms of violence lead to the nurse seeking and successfully finding increased outside support, and/or they indicate a survivor artifact, whereby nurses with strong outside work social support remain at the workplace, whereas those without such support leave the organization. Furthermore, this result may indicate a contamination effect, whereby workplace-based social support is not beneficial because of being based in the context where the violence occurred.
Regarding NA, this individual variable was related to internal and external emotional abuse, as well as external threat of assault. The pattern of findings suggests that NA may play a role where the violence is of a more general nature and internally processed and appraised. This finding may represent the perception mechanism of NA,18,20 where high-NA people give more attention to potential threats and perceive ambiguous stimuli negatively.26 Therefore, the study’s results extend prior research on the centrality of NA’s role in workplace bullying18,25 to forms of violence.
Taken together, the results of this study provide indications of where work may be designed to structure out violence, while also considering the inherent nature of perceptions, especially NA. Elements of the JD-R model and NA were shown to be useful in identifying relationships for workplace violence, not just bullying acts. The relationships between job demands and each of the externally sourced forms of violence highlight how workloads and busyness are central to the nature of the interaction with patients and their family and friends. Higher levels of job control enabled nurses to better avoid certain violent incidents, whereas higher levels of social support from outside work may have been a resource used by nurses who had suffered physical assault or sexual harassment to continue at work. Furthermore, the findings surrounding NA and particular forms of workplace violence highlighted the internal processes and appraisals.
In response to calls for increased involvement of nurse leaders to be involved in the prevention and intervention of workplace violence,32 these results highlight key practical implications for nurses who care for the elderly, whereby demanding and restrictive procedures place these nurses at greater risk of workplace violence. For instance, changes to scheduling may help to reduce this risk. Allowing the reallocation of work tasks to avoid busy patient visiting times (eg, lunchtime and after standard work hours) may decrease the amount of exposure to family members and friends and provide an opportunity for nurses to exercise autonomy over their work. Furthermore, the individual variation associated with NA presents a challenge for nurse managers, where perceptions influence nurses’ reality for certain forms of violence. Diversity training for nurse unit managers that focuses on these potential sensitivities and how certain types of violence can involve subjective components may be of benefit.
The main limitation of this study is that it is based on a cross-sectional data. Subsequently, the data do not allow the causal directions between variables to be established. The generalizability of the results may also be limited as only 1 healthcare organization from Australia was sampled. Although the response rate of the study was high, and the demographic characteristics were representative of the Australian nurses who care for elderly populations, future research that examines other healthcare organizations from other countries is necessary to replicate the current findings and provide comparative international data. Moreover, the results are limited in transferability across nursing contexts outside the care of the elderly. It may be beneficial for future research to investigate whether similar or dissimilar factors are involved in workplace violence experienced by other types of nursing staff. Finally, in regard to the workplace violence measure adopted,20 the response option requested that nurses indicate their level of exposure in the past 5 work shifts, potentially limiting our understanding of ongoing repeated exposure. This response option was in line with the actual measure used by Hesketh et al,20 who indicated that such a response ensured accurate recall. Future researchers could possibly add another response option with an increased time period to gather more information about this level of exposure.
This study found high levels of workplace violence in the context of zero tolerance policies and extends previous research on workplace bullying and nurses in general to workplace violence in a settings providing care for the elderly. In the context of an aging population, elderly care demand will continue to increase,14 and if this demand translates to the job level, the relationship between job demands and the external forms of violence analyzed in this study implies that violence against nurses caring for the elderly is likely to increase. This study’s aim of understanding antecedents of violence can inform future research and nurse manager actions about the most appropriate intervention and prevention approaches, particularly regarding designing work to structure out violence, bearing in mind the subjective nature of some forms of workplace violence.
1. Claravall L. Healthcare violence: a nursing administration perspective. J Nurs Adm. 1996; 26 (2): 41–46.
2. Autrey PS, Howard JL, Wech BA. Sources, reactions, and tactics used by RNs to address aggression in an acute care hospital. J Nurs Adm. 2013; 43 (3): 155–159.
3. Hogh A, Viitasara E. A systematic review of longitudinal studies of nonfatal workplace violence. Eur J Work Organ Psychol. 2005; 14 (3): 291–313.
4. Nabb D. Visitors’ violence: the serious effects of aggression on nurses and others. Nurs Stand. 2000; 14 (23): 36–38.
5. Rowe MM, Sherlock H. Stress and verbal abuse in nursing: do burned out nurses eat their young? J Nurs Manag. 2005; 13 (3): 242–248.
6. Hutchinson M, Jackson D, Wikes L, Vickers MH. A new model of bullying in the nursing workplace: organizational characteristics as critical antecedents. Adv Nurs Sci. 2008; 31 (2): 60–71.
7. Hutchinson M, Wikes L, Jackson D, Vickers MH. Integrating individual, work group and organizational factors: testing a multidimensional model of bullying in the nursing workplace. J Nurs Manag. 2010; 18 (2): 173–181.
8. Quine L. Workplace bullying in nurses. J Health Psychol. 2001; 6 (1): 73–84.
9. Simons S. Workplace bullying experienced by Massachusetts registered nurses and the relationship to intention to leave the organization. Adv Nurs Sci. 2008; 31 (2): E48–E59.
10. Gacki-Smith J, Juarez AM, Boyett L, Homeyer C, Robinson L, MacLean SL. Violence against nurses working in US emergency departments. J Nurs Adm. 2009; 39 (7/8): 340–349.
11. Eley R, Hegney D, Buikstra E, Fallon T, Plank A, Parker V. Aged care nursing in Queensland—the nurses’ view. J Clin Nurs. 2007; 16 (5): 860–872.
12. Estryn-Behar M, van der Heijden B, Camerino D, Fry C, Le Nezet O, Conway PM, Hasselhorn HM. Violence risks in nursing—results from the European ‘NEXT’ study. Occup Med. 2008; 58: 107–114.
13. Hegney D, Eley R, Plank A, Buikstra E, Parker V. Workplace violence in Queensland, Australia: the results of a comparative study. Int J Nurs Pract. 2006; 12: 220–231.
14. Access Economics. Nurses in Residential Aged Care. Australia: Access Economics; 2009.
15. Baillien E, Rodríguez-Muñoz A, de Witte H, Notelaers G, Moreno-Jiménez B. The Demand-Control model and target’s reports of bullying at work: a test within Spanish and Belgian blue-collar workers. Eur J Work Organ Psychol. 2011; 20: 157–177.
16. Notelaers G, Baillien E, de Witte H, Einarsen S, Vermunt JK. Testing the strain hypothesis of the Demand Control Model to explain severe bullying at work. Econ Ind Democracy. 2012; 34 (1): 69–87.
17. Tuckey MR, Dollard MF, Hosking PJ, Winefield AH. Workplace bullying: the role of psychosocial work environment factors. Int J Stress Manage. 2009; 16 (3): 215–232.
18. Mikkelsen EG, Einarsen S. Relationships between exposure to bullying at work and psychological and psychosomatic health complaints: the role of state negative affectivity and generalized self-efficacy. Scand J Psychol. 2002; 43: 397–405.
19. Spector PE, Zapf D, Chen PY, Frese M. Why negative affectivity should not be controlled in job stress research: don’t throw out the baby with the bath water. J Organ Behav. 2000; 21 (1): 79–95.
20. Hesketh KL, Duncan SM, Estabrooks CA, et al. Workplace violence in Alberta and British Columbia hospitals. Health Policy. 2003; 63: 311–321.
21. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The Job Demands-Resources model of burnout. J Appl Psychol. 2001; 86 (3): 499–512.
22. Schaufeli WB, Bakker AB. Job demands, job resources, and their relationship with burnout and engagement: a multi-sample study. J Organ Behav. 2004; 25 (3): 293–315.
23. Skogstad A, Torsheim T, Einarsen S, Hauge LJ. Testing the work environment hypothesis of bullying on a group level of analysis: psychosocial factors as precursors of observed workplace bullying. Appl Psychol. 2011; 60 (3): 475–495.
24. Baillien E, Neyens I, de Witte H, de Cuyper N. A qualitative study on the development of workplace bullying: towards a three way model. J Community Appl Soc Psychol. 2009; 19 (1): 1–16.
25. Hansen ÅM, Hogh A, Persson R, Karlson B, Garde AH, Ørbæk P. Bullying at work, health outcomes, and physiological stress response. J Psychosom Res. 2006; 60 (1): 63–72.
26. Watson D, Clark LA. Negative affectivity: the disposition to experience aversive emotional states. Psychol Bull. 1984; 96 (3): 465–490.
27. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988; 54: 1063–1070.
28. Caplan R, Cobb S, French J Jr, Harrison R, Pinneau S. Job Demands and Worker Health: Main Effects and Occupational Differences. Ann Arbor, MI: The Institute for Social Research; 1980.
29. Karasek R. Job Content Questionnaire and User’s Guide. Los Angeles, CA: Department of Industrial and Systems Engineering; 1985.
30. Podsakoff PM, Organ DW. Self-reports in organizational research: problems and prospects. J Manag. 1986; 12 (4): 531.
31. Faul FEE, Lang AG, Buchner A. G*Power 3: a flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods. 2007; 39 (2): 175–191.
32. Hardin D. Strategies for nurse leaders to address aggressive and violent events. J Nurs Adm. 2012; 42 (1): 5–8.