ARTICLES

Organizational Justice and Collaboration Among

Nurses as Correlates of Violent Assaults by Patients
in Psychiatric Care
Virve Maaret Pekurinen, R.N., M.N.Sc., Maritta Välimäki, R.N., Ph.D., Marianna Virtanen, Ph.D., Paula Salo, Ph.D.,
Mika Kivimäki, Ph.D., Jussi Vahtera, M.D., Ph.D.

Objectives: This study tested the hypothesis that poor or- were not dependent to a significant degree on organiza-
ganizational justice and collaboration among nurses are as- tional justice, nor were patients’ assaults dependent on
sociated with increased stress among nurses, which, in turn, stress levels. However, low organizational justice and poor
increases the likelihood of violent assaults by patients. collaboration between nurses were associated with in-
creased reports of violent assaults by patients in psychiatric in-
Methods: A cross-sectional survey was conducted of nurses patient settings (p,.05 for both). The model explained 5.7%
in 90 psychiatric inpatient wards in five hospital districts and of violent assaults at nearly significant levels (p=.052).
one regional hospital in Finland. A total of 758 nurses (reg-
istered nurses or enrolled/mental health nurses) responded Conclusions: These findings suggest that organizational
to the survey. Self-administered postal questionnaires were justice, collaboration between staff members, and violent
used to assess organizational justice, collaboration, nurses’ assaults by patients are linked in psychiatric inpatient set-
stress, and violent assaults by patients. Structural equation tings. Evaluating a variety of factors, including issues related
modeling (SEM) was used in model testing. to organizational justice and collaboration among nurses,
may be useful to minimize assaults by patients in psychiatric
Results: SEM did not support a role for stress in mediating settings.
between organizational justice, collaboration between nurses,
and violent assaults by patients, given that stress levels Psychiatric Services 2017; 68:490–496; doi: 10.1176/appi.ps.201600171

Factors predicting violent assaults on staff by patients in psy- and staff may arise, which may result in patient assaults (14).
chiatric inpatient settings are poorly understood. Some studies This observation supports earlier findings that have associ-
suggest that the risk of committing assaults is increased among ated high job strain, psychological distress (15), job demands
patients with certain diagnoses, such as schizophrenia (1,2), (16), time pressure at work (9), and problems in staff-patient
affective disorder (3), and impulse control disorder (4). Other interaction (17,18) with patient assaults. Likewise, workplace
patient characteristics, such as being male (2), having a his- support (15), interpersonal relationships between staff (8),
tory of violence (3,5) or substance abuse (2,3), having sleeping quality of teamwork (9), and organizational justice (15) may
problems (6) or having poor self-reflective skills (5), have also play a role in patient assaults.
been associated with assaults. Justice refers to an action or decision that is understood
However, patient characteristics may explain only a pro- to be morally right on the basis of ethics, religion, fairness,
portion of violent assaults (7). Other factors may increase the equity, or law (19). Organizational justice, originally derived
risk of being subjected to violent assaults, such as shift work from equity theory (20), refers to an employee’s perception of
(8,9) and a fixed schedule of night work (10), poor informa- his or her organization’s behaviors, decisions, and actions and
tion flow among coworkers (8), patient overcrowding (11), and how these influence the employee’s own attitudes and be-
uncertainty among nurses regarding treatment (9). Nurses’ haviors at work (21). Previous research has shown that low
characteristics, such as being male (8,9), being young (8,9,12), organizational justice causes increases in employees’ stress
having a lower level of qualifications (9,12), or having less levels (22,23), intragroup conflicts (24) and work group mis-
training (12) or work experience (12,13), may also be associ- behavior (25). Poor teamwork creates inadequate program
ated with an increased risk of being subjected to assaults. One organization, which results in higher levels of stress among
qualitative study reported that when nurses feel pressured at nurses (26) and may cause additional assaults on psychiatric
work, distractions or miscommunications between patients wards (7,27).

490 ps.psychiatryonline.org Psychiatric Services 68:5, May 2017

 PEKURINEN ET AL.

Despite inconsistency in the literature about the causes of Respondents also indicated the month in which the exposure
patient violence, the topic has attracted constant attention occurred, from 1, January, to 12, December. The occurrence of
from researchers (8,28). It has been suggested as one of the violent assaults by patients was combined into a sum score by
main reasons for decreased commitment (8,9) to an organi- calculating the number of months in which any of the four
zation among staff, for the intention of leaving the profession types of violence had occurred during the past 12 months
(9,29), as well as for accidents, disability, death, absenteeism (range 0–48). In this study, the internal consistency of the scale
(29), negative feelings (30), lower job satisfaction (29) and was respectable (.77), as measured by the Kuder-Richardson
burnout (9) among staff members. Given that patient violence formula.
toward nurses in psychiatric settings is a complex and multi- Organizational justice was measured by using a ques-
dimensional problem (7), there is an urgent need to identify tionnaire of procedural and relational justice adopted from
the factors contributing to its prevalence. Moorman’s organizational justice measure (31,32). Pro-
In this cross-sectional survey study, we developed a hy- cedural justice refers to the extent that decision-making
pothesis that extended the ideas from existing research on procedures include input from all parties affected, are con-
the interaction of organizational justice, collaboration be- sistently applied, are accurate, suppress bias, and are cor-
tween nurses, nurses’ stress, and patient violence. On the rectable and ethical. Relational justice refers to considerate,
basis of this hypothesis, we formulated and tested a model polite, and fair treatment of individuals (33). The question-
using the following assumptions. First, the perception of low naire measures respondents’ current opinions on procedural
organizational justice by nurses is associated with increased justice (seven items) and relational justice (six items) at their
stress, which in turn is associated with an increased number organization on a 5-point scale, from 1, totally disagree, to 5,
of violent assaults by patients. Second, low organizational totally agree, with higher scores indicating better organiza-
justice is associated with poor collaboration among nurses. tional justice. A mean scale score was calculated for both
Third, poor collaboration among nurses is associated with scales by averaging the scores on each item. The instrument
increased stress, which in turn is associated with increased has been used with Finnish health care staff (34), and its
numbers of violent assaults by patients. internal consistency has been strong (procedural justice,
a=.90 [32], a=.80 [35]; relational justice, a=.81 [32], a=.90
[35]). In our data, the internal consistency of the scales
METHODS
remained strong (procedural justice, a=.94; relational jus-
Participants and Procedures tice, a=.91).
Participants were selected from the Finnish Public Sector Collaboration was measured by using two subscales
(FPS) study cohort, which includes employees in ten towns derived from the 14-item Team Climate Inventory (TCI)
and six hospital districts. Employers’ records are used to (36,37). Participative safety, with four items, measures the
identify eligible employees for nested survey cohorts that extent to which “involvement in decision-making is moti-
have been sent questionnaires by mail or e-mail every four vated and reinforced while occurring in an environment
years since 2000. For our study, we used a subset of FPS which is perceived as interpersonally nonthreatening.”
cross-sectional questionnaire data collected in 2012 from Support for innovation, with three items, refers to the “ex-
five of the nation’s 20 hospital districts and one regional pectation, approval, and practical support of attempts to
hospital providing specialized psychiatric care. Eligible introduce new and improved ways of doing things in the
participants were nurses (registered nurses and licensed work environment” (38). Items are rated from 1, totally
practical nurses) working on the 90 psychiatric inpatient disagree, to 5, totally agree, with higher scores indicating
wards operational at the time of the survey (N=1,033). Of better collaboration. A mean scale score was calculated by
these, 758 (73%) responded to the survey in Finnish mea- averaging the scores on each item. The subscales have been
suring psychosocial work environment and patient assaults, used with Finnish health care staff (39). The internal con-
which was part of the FPS questionnaire survey exploring sistency of the subscales has been strong in earlier studies
behavioral and psychosocial factors and health. The Ethics (participative safety, a=.87; support for innovation, a=.81
Committee of the Helsinki and Uusimaa Hospital District [40]) and remained strong in our data (participative safety,
approved the study. The principles of the Declaration of a=.86, support for innovation, a=.82).
Helsinki were followed. Nurses’ psychological distress, or stress, was measured
with the 12-item General Health Questionnaire (GHQ-12),
Measures which measures minor psychiatric morbidity (41). Respon-
The occurrence of violent assaults by patients was surveyed dents rate the extent to which they have experienced the
retrospectively with a measure developed for the purposes symptoms of distress in the past few weeks, from 0, not at all,
of the FPS study (11). Respondents are asked whether they to 3, much more than usual, with higher scores indicating
had encountered any of the four listed types of violent in- greater stress. A mean scale score was calculated by averaging
cidents at work (verbal threats; physical violence, such as the scores on each item.
hitting or kicking; assaults on ward property, such as throwing The scale has previously been used as an indicator of
objects; and armed threats during the past year (yes=1, no=0). stress (42–45). The GHQ-12 has been used with Finnish

Psychiatric Services 68:5, May 2017 ps.psychiatryonline.org 491

CORRELATES OF VIOLENT ASSAULTS BY PATIENTS IN PSYCHIATRIC CARE

FIGURE 1. The role of stress (STR) as a mediator of the effect of organizational included nonsignificant chi-square statistics as
justice (OJ) and collaboration (COLL) among nurses on violent assaults (VA) by well as findings for the comparative fit index
patients (model 1) and the effect of omitting stress from the model (model 2)a
(CFI), the Tucker-Lewis Index (TLI), the
standardized root-mean-square residual
(SRMR), and the root-mean-square error of
approximation (RMSEA). The chi-square test
is an absolute test of model fit, so the model is
rejected in case of p,.05. CFI values may range
between 0 and 1, with values close to 1 indi-
cating very good fit (50); in this study the CFI
was set at ..95. Further, a TLI index close to 1.0
and RMSEA values ,.05 were set as criteria for
a fit model (51). SRMR, the most sensitive index
for detecting misspecified latent structures or
factor covariances, was set at #.08 (51). The
model’s ability to explain assaults was assessed
by using the coefficient of determination (R2)
(52). Mplus was used for the SEM, and SPSS,
version 21, was used for the other analyses.

RESULTS
Descriptive Characteristics
The majority of participants were female
(74%) registered nurses (58%) who worked
full-time (95%) on a permanent employment
contract (78%). The majority had been ex-
posed to verbal threats (59%, N=424) dur-
ing the past year; 46% (N=338) reported
assaults on ward property, 35% (N=251) re-
ported exposure to physical violence, and
a
RJ, relational justice; PJ, procedural justice; PS, participative safety; SI, support for in- 5% (N=34) reported receiving armed threats.
novation. Latent variables are depicted as ovals, and observed variables are depicted as
rectangles. Demographic and work-related information
about the participants is presented in Table 1.
health care staff (46) and has been validated in the Finnish Table 2 presents the mean6SD scores and internal con-
population (47). The internal consistency of the scale has been sistency values for each observed variable (participative safety,
strong (a=.90 [48], a=.85 [49]), and it remained strong in our support for innovation, relational justice, procedural justice,
data (a=.88). nurses’ psychological distress, and violent assaults by patients)
All instruments (organizational justice scale, the TCI, as well as correlations between the observed variables.
and the GHQ-12), which were originally written in English,
had been translated to Finnish before this study. Constructed Structural Equation Models
In the original model (model 1), stress was considered a
Data Analysis mediator between organizational justice, collaboration,
Our proposed model consisted of organizational justice, and patients’ assaults (Figure 1). That model was rejected
collaboration among nurses, stress, and patient violent as- because of poor model fit, indicated by significant chi-
saults. The model construction is described in model 1 of square values and RMSEA values (90% confidence interval
Figure 1. Stress was considered as a mediator between the two [CI]=.03–.08). The role of stress as a mediating factor was
factors (organizational justice and collaboration) and violent also rejected because stress levels were not dependent to a
assaults by patients. The model was encoded into a multiple significant degree on organizational justice, nor were pa-
regression equation by arrows indicating the relationships tients’ assaults dependent on stress levels. Therefore, the
between specific factors. The fit of the model was determined explanation of assaults in model 1 did not reach statistical
by using structural equation modeling (SEM) with maximum significance.
likelihood estimations to test the hypothesized model. SEM Based on these parameter estimates, we modified the
was chosen because it is suitable for confirmatory testing model by removing the mediating factor of stress to achieve
of hypothesized models that are supported by either theories better goodness of fit (model 2 in Figure 1). Akaike’s in-
or empirical research. Criteria for goodness of fit of the model formation criterion (AIC) and the Bayesian information

492 ps.psychiatryonline.org Psychiatric Services 68:5, May 2017

 PEKURINEN ET AL.

criterion (BIC) were used to compare the alternative models TABLE 1. Demographic characteristics and work-related
(53). The overall lowest values of AIC and BIC represent the information for 758 nurses at psychiatric inpatient wards in
Finland
best model fit (54). The results from the analysis of model
2 indicated a more acceptable model fit on all indices Characteristic N %
compared with model 1 (RMSEA, CI=.00–.05). AIC and BIC Age (M6SD) 43.96610.95
indices were lower in model 2 compared with model 1, also Gender
indicating a better fit for model 2. Furthermore, in model 2, Female 558 74
Male 200 26
relationships between factors were all statistically signifi-
Marital status
cant at the .05 level. Organizational justice was positively
Married or cohabiting 568 75
related to collaboration among nurses, suggesting that low Divorced or separated 89 12
organizational justice is associated with poor collaboration Single 92 12
among nurses. Organizational justice was negatively related Widowed 7 1
to assaults, suggesting that lower organizational justice is Professional status
associated with more frequent assaults. Collaboration was Registered nurse 436 58
positively related to assaults, which may indicate that better Licensed practical nurse 241 32
Head nurse 81 10
collaboration among nurses is associated with more frequent
Employment
assaults. However, the correlations between the observed
Permanent 592 78
variables related to collaboration (participative safety and Fixed term 166 22
support for innovation) and assaults were negative (Table 2), Nature of job
indicating a negative relationship between collaboration and Full-time 722 95
assaults. The association between collaboration and assaults Part-time 36 5
might be affected by the strong associations between orga- Time of work
nizational justice and collaboration factors (p#.001) and Day 235 30
between organizational justice factors and assaults (p=.001), Shift work without nights 118 16
Shift work with nights 367 49
which could create a false-positive dependency. Therefore,
Night 28 4
we may assume that the relationship was negative, rather Other irregular work 9 1
than positive, indicating that poor collaboration among nurses Duration of employment (M6SD years)
was associated with more frequent patient assaults. Model Current organization 9.1468.78
2 explained 5.7% of patient assaults at nearly significant Current position 8.0168.67
levels (p=.052).
Table 3 shows the goodness-of-fit indices and the co-
efficient of determination (R2) for the alternative models for The very nature of the work performed by psychiatric
explaining violent assaults by patients. nurses may explain the contradictory study results. For
example, one study reported that nurses’ mental health
status, as measured by the GHQ-12, was not associated with
DISCUSSION
patient violence in psychiatric settings, whereas such an
To examine violent assaults by patients on psychiatric association was found in other settings (29). There may be
wards, we hypothesized that nurses’ stress was a mediator several reasons for this discrepancy. It can be assumed, for
between other model factors (organizational justice and example, that psychiatric nurses are more accustomed to
collaboration among nurses) and patient violent assaults, dealing with aggressive patients compared with nurses in
and we developed a model to test that hypothesis. However, other medical fields. Also, the behavior of psychiatric nurses
stress was not related either to violent assaults by patients or may not be as strongly affected by stress compared with that
to organizational justice, and therefore the mediating role of of nurses working in other specialties.
stress was not supported. It is also possible that the instrument used in this study
Although we are unaware of studies that are highly sim- did not capture the dimensions of stress that have been
ilar to ours, we assume that our results, surprisingly, are not previously documented to be associated with violence. For
likely to be in line with those of earlier studies. For example, example, the Italian cross-sectional study found certain as-
in a cross-sectional study conducted among workers in the pects of stress, such as job demands and poor workplace
Italian public health care sector, indications were found that social support, as defined in Karasek’s model (55), to be risk
psychological disorders among staff, measured by the same factors for violence (15). These types of stress—increased job
questionnaire as used in our study, preceded certain types of demands (15) and pressures (14) and lack of support in the
violence toward staff (15). However, the study population, workplace (15)—were not captured by the measure of psy-
consisting of all professionals working in any specialty in the chological distress used in our study.
public health care sector, differed greatly from our study Our results regarding the association of poor collabora-
population, comprising only nurses working on psychiatric tion among nurses and patient violence are in line not only
wards. with those of the Italian cross-sectional study concerning

Psychiatric Services 68:5, May 2017 ps.psychiatryonline.org 493

CORRELATES OF VIOLENT ASSAULTS BY PATIENTS IN PSYCHIATRIC CARE

TABLE 2. Correlations between scores on observed variables among 758 nurses longitudinal research is needed to evaluate the
at psychiatric inpatient wards in Finland impact of organizational justice and collabo-
Score ration on patient assaults.
Variable M SD a PS SI RJ PJ STR VA Second, relying on nurses’ retrospective
PS a
3.7 .80 .86 —
recall of assaults may have caused some
SIa 3.3 .82 .82 .606 — misclassifications. Staff may overestimate the
RJa 3.7 .98 .91 .477 .441 — frequency of assaults, for example, although
PJa 2.9 .81 .94 .352 .358 .416 — other assessment methods—such as daily staff
STRb 1.9 .39 .88 –.186 –.197 –.100 –.141 — reports, standard instruments, and official
c d
VA 4.6 8.37 .77 –.056 –.024 –.134 –.108 –.018 —
incident reports (56–58)—may underreport
a
Possible scores for participative safety (PS), support for innovation (SI), relational justice (RJ), and assaults, irrespective of the severity of the
procedural justice (PJ) range from 1 to 5, with higher scores indicating better collaboration
among nurses (PS and SI) and better organizational justice (RJ and PJ). assault (56). Staff may consider assaults part
b
Possible scores for nurses’ psychological distress (STR) in the past few weeks range from 0, not of their job (59) or feel embarrassed about
at all, to 3, much more than usual. being assaulted (60), which may increase
c
Possible scores range from 0 to 48, indicating the number of months in which a nurse expe-
rienced each of the four types of violent assaults (VA) by patients (verbal threats; physical underreporting. It has been suggested that
violence, such as hitting or kicking; assaults on ward property, such as throwing objects; and self-reporting methods that rely on memory,
armed threats) during the past 12 months. like other types of assessment methods, are
d
Internal consistency of the observed variables was measured by the Kuder-Richardson formula.
likely to underestimate the occurrence of
assaults (61). However, the validity of our
workplace support (15) but also with other findings (8,9). measurement for assessing the occurrence of assaults is
Quality of teamwork (9) and workplace interpersonal rela- supported by earlier studies that have found an increasing
tionships (8) have also been associated with violence. Good risk of self-reported physical assaults connected to patient
collaboration among nurses may have a positive effect on the overcrowding, a risk of violence in psychiatric settings (11),
team’s ability to respond to violence and may add to an and an exceptionally high risk of exposure to mental abuse
overall atmosphere of calm on the ward, which may reduce and physical violence among special education teachers
patient aggression. compared with their colleagues in general education (62). In
Our findings regarding an association between lower addition, the occurrence of aggression found in this study
perceptions of organizational justice by nurses and in- is quite similar to the findings of earlier studies (28,63).
creased patient assaults are in accordance with those of an Third, the model explained only a small amount of the
earlier study (15). However, the mechanisms remain un- variance in patient assaults, which might raise questions
known. Research has shown that perceptions of low jus- about the significance of the findings. However, we had no
tice negatively affect workers’ behavior in groups (25) and information on the most important predictors of aggression,
increase intragroup conflicts among nurses (24). There- such as patient characteristics or severity of the disease.
fore, we could draw the tentative conclusion that low Thus it is to be expected that the model would explain a
justice perceptions not only may negatively affect nurses’ small amount of variance in patient assaults. It should be
behavior toward colleagues but also may contribute to noted that the associations between model factors were
poor staff-patient interactions and alter nurses’ behavior statistically significant. Thus the study contributed to the
toward patients, which may be associated with increased understanding of the phenomenon of patient violence to-
patient assaults (18). ward psychiatric nursing staff, even though its purpose was
This study had limitations. First, the cross-sectional design not to make precise predictions about the role of various
prevents us from making causal statements about the results. factors in assaults by patients.
The fact that patient assaults were evaluated retrospectively,
whereas other model variables were based on nurses’ current
CONCLUSIONS
experiences, may have resulted in a reversal of the direction
of causality proposed in the hypothesized model (for exam- Nurses’ perceptions of poor organizational justice and poor
ple, increased assaults may predict poor collaboration and collaboration among nurses were found to be linked to in-
low organizational justice rather than vice versa). Therefore, creased patient assaults, whereas nurses’ stress, as measured
by psychological distress, was
TABLE 3. Goodness-of-fit indices for alternative models of mechanisms involved in violent assaults not linked to increased pa-
by patients on nurses in psychiatric inpatient wardsa tient assaults. Longitudinal
Model x 2
p df CFI TLI AIC BIC SRMR RMSEA R 2b
p research is needed to verify
our findings and determine
1 23.66 .001 7 .979 .955 12,940.814 13,033.427 .037 .056 .000 .800
2 1.82 .611 3 1.000 1.005 12,217.929 12,296.651 .007 .000 .057 .052 the direction of causality. Also,
a
future research should attempt
CFI, comparative fit index; TLI, Tucker-Lewis Index; AIC, Akaike’s information criterion; BIC, Bayesian information
criterion; SRMR, standardized root-mean-square residual; RMSEA, root-mean-square error of approximation to clarify the mechanisms
b
The coefficient of determination indicates the model’s ability to explain violent assaults by patients. underlying the associations

494 ps.psychiatryonline.org Psychiatric Services 68:5, May 2017

 PEKURINEN ET AL.

between nurses’ work-related stress and patient assaults in 10. Nabe-Nielsen K, Tüchsen F, Christensen KB, et al: Differences
the context of psychiatric nursing, especially the aspects of between day and nonday workers in exposure to physical and
psychosocial work factors in the Danish eldercare sector. Scandi-
stress that may increase the risk of assaults. In addition, the
navian Journal of Work, Environment and Health 35:48–55, 2009
mechanisms underlying the association between nurses’ 11. Virtanen M, Vahtera J, Batty GD, et al: Overcrowding in psychi-
perceptions of organizational justice and patient aggression atric wards and physical assaults on staff: data-linked longitudinal
must be clarified. study. British Journal of Psychiatry 198:149–155, 2011
Our findings suggest that evaluating a variety of factors, 12. Flannery RB Jr, LeVitre V, Rego S, et al: Characteristics of staff
victims of psychiatric patient assaults: 20-year analysis of the
including organizational justice and collaboration-related is-
Assaulted Staff Action Program. Psychiatric Quarterly 82:11–21,
sues, both on the frontline and at the administrative level, is 2011
important in minimizing patient assaults in psychiatric settings. 13. Privitera M, Weisman R, Cerulli C, et al: Violence toward mental
health staff and safety in the work environment. Occupational
AUTHOR AND ARTICLE INFORMATION Medicine 55:480–486, 2005
14. Ward L: Ready, aim fire! Mental health nurses under siege in acute
Ms. Pekurinen is with the Department of Nursing Science, Faculty of
inpatient facilities. Issues in Mental Health Nursing 34:281–287,
Medicine, Dr. Salo is with the Department of Psychology, and Dr. Vahtera is
2013
with the Department of Public Health, Faculty of Medicine, all at the
15. Magnavita N, Heponiemi T: Violence towards health care workers
University of Turku, Turku, Finland (e-mail: vimapek@utu.fi). Dr. Salo is also
in a public health care facility in Italy: a repeated cross-sectional
with the Finnish Institute of Occupational Health, Turku. Dr. Vahtera is also
study. BMC Health Services Research, 2012 (doi 10.1186/1472-
with Turku University Hospital, where Dr. Välimäki is affiliated. Dr. Välimäki
6963-12-108)
is also with the School of Nursing, Hong Kong Polytechnic University,
16. Demir D, Rodwell J: Psychosocial antecedents and consequences
Hong Kong, China (SAR). Dr. Virtanen is in the Helsinki office of the Finnish
of workplace aggression for hospital nurses. Journal of Nursing
Institute of Occupational Health. Dr. Kivimäki is with Department of Epi-
Scholarship 44:376–384, 2012
demiology and Public Health, University College London and with Clin-
17. Wenzhi Cai, Ling Deng, Meng Liu, et al: Antecedents of medical
icum, Faculty of Medicine, University of Helsinki, Helsinki.
workplace violence in South China. Journal of Interpersonal Vi-
This study was funded by the Finnish Work Environment Fund (111298). olence 26:312–327, 2011
The Finnish Public Sector study was supported by the Academy of 18. Papadopoulos C, Ross J, Stewart D, et al: The antecedents of vi-
Finland (Projects 264944 and 267727), the Finnish Work Environment olence and aggression within psychiatric inpatient settings. Acta
Fund (115421) and the participating organizations. The authors thank Psychiatrica Scandinavica 125:425–439, 2012
Jaana Pentti, B.Sc., for preparing the data for analysis and Jouko 19. Tabibnia G, Satpute AB, Lieberman MD: The sunny side of fair-
Katajisto, M.Sc., for conducting the data analysis. ness: preference for fairness activates reward circuitry (and dis-
The authors report no financial relationships with commercial interests. regarding unfairness activates self-control circuitry). Psychological
Science 19:339–347, 2008
Received April 9, 2016; revision received September 1, 2016; accepted
20. Adams JS: Inequity in social exchange; in Advances in Experi-
October 7, 2016; published online February 1, 2017.
mental Social Psychology, vol 2. Edited by Berkowitz L. New York,
Academic Press, 1965
REFERENCES 21. Greenberg J: A taxonomy of organizational justice theories. Acad-
1. Amoo G, Fatoye FO: Aggressive behaviour and mental illness: a emy of Management Review 12:9–22, 1987
study of inpatients at Aro Neuropsychiatric Hospital, Abeokuta. 22. Sutinen R, Kivimäki M, Elovainio M, et al: Organizational fairness
Nigerian Journal of Clinical Practice 13:351–355, 2010 and psychological distress in hospital physicians. Scandinavian
2. Dack C, Ross J, Papadopoulos C, et al: A review and meta-analysis Journal of Public Health 30:209–215, 2002
of the patient factors associated with psychiatric in-patient ag- 23. Dackert I: The impact of team climate for innovation on well-being
gression. Acta Psychiatrica Scandinavica 127:255–268, 2013 and stress in elderly care. Journal of Nursing Management 18:
3. Stewart D, Bowers L: Inpatient verbal aggression: content, targets 302–310, 2010
and patient characteristics. Journal of Psychiatric and Mental 24. Almost J, Doran DM, McGillis Hall L, et al: Antecedents and
Health Nursing 20:236–243, 2013 consequences of intra-group conflict among nurses. Journal of
4. Sadock BJ, Sadock VA, Ruiz P, et al: Kaplan and Sadock’s Com- Nursing Management 18:981–992, 2010
prehensive Textbook of Psychiatry. Philadelphia, Wolters Kluwer/ 25. Priesemuth M, Arnaud A, Schminke M: Bad behavior in groups:
Lippincott Williams & Wilkins, 2009 the impact of overall justice climate and functional dependence on
5. Ekinci O, Ekinci A: Association between insight, cognitive insight, counterproductive work behavior in work units. Group and Or-
positive symptoms and violence in patients with schizophrenia. ganization Management 38:230–257, 2013
Nordic Journal of Psychiatry 67:116–123, 2013 26. Bowers L, Nijman H, Simpson A, et al: The relationship between
6. Kamphuis J, Dijk DJ, Spreen M, et al: The relation between poor leadership, teamworking, structure, burnout and attitude to pa-
sleep, impulsivity and aggression in forensic psychiatric patients. tients on acute psychiatric wards. Social Psychiatry and Psychi-
Physiology and Behavior 123:168–173, 2014 atric Epidemiology 46:143–148, 2011
7. Cutcliffe JR, Riahi S: Systemic perspective of violence and ag- 27. Whittington R, Wykes T: An observational study of associa-
gression in mental health care: towards a more comprehensive tions between nurse behaviour and violence in psychiatric hos-
understanding and conceptualization: part 2. International Journal pitals. Journal of Psychiatric and Mental Health Nursing 1:85–92,
of Mental Health Nursing 22:568–578, 2013 1994
8. Camerino D, Estryn-Behar M, Conway PM, et al: Work-related 28. Spector PE, Zhou ZE, Che XX: Nurse exposure to physical and
factors and violence among nursing staff in the European NEXT nonphysical violence, bullying, and sexual harassment: a quanti-
study: a longitudinal cohort study. International Journal of Nurs- tative review. International Journal of Nursing Studies 51:72–84,
ing Studies 45:35–50, 2008 2014
9. Estryn-Behar M, van der Heijden B, Camerino D, et al: Violence 29. Merecz D, Rymaszewska J, Moscicka A, et al: Violence at the
risks in nursing—results from the European “NEXT” Study. Oc- workplace—a questionnaire survey of nurses. European Psychiatry
cupational Medicine 58:107–114, 2008 21:442–450, 2006

Psychiatric Services 68:5, May 2017 ps.psychiatryonline.org 495

CORRELATES OF VIOLENT ASSAULTS BY PATIENTS IN PSYCHIATRIC CARE

30. Needham I, Abderhalden C, Halfens RJ, et al: Nonsomatic effects 48. Virtanen M, Kivimäki M, Elovainio M, et al: From insecure to
of patient aggression on nurses: a systematic review. Journal of secure employment: changes in work, health, health related be-
Advanced Nursing 49:283–296, 2005 haviours, and sickness absence. Occupational and Environmental
31. Moorman R: Relationship between organizational justice and Medicine 60:948–953, 2003
organizational citizenship behaviors: do fairness perceptions in- 49. Ip WY, Martin CR: Psychometric properties of the 12-item General
fluence employee citizenship? Journal of Applied Psychology 76: Health Questionnaire (GHQ-12) in Chinese women during preg-
845–855, 1991 nancy and in the postnatal period. Psychology Health and Medi-
32. Elovainio M, Kivimäki M, Vahtera J: Organizational justice: evi- cine 11:60–69, 2006
dence of a new psychosocial predictor of health. American Journal 50. Bentler PM: Comparative fit indexes in structural models. Psy-
of Public Health 92:105–108, 2002 chological Bulletin 107:238–246, 1990
33. Kramer RM, Tyler TR: Trust in Organizations: Frontiers of Theory 51. Hu L, Bentler P: Cutoff criteria for fit indexes in covariance
and Research. London, Sage, 1996 structure analysis: conventional criteria versus new alternative.
34. Heponiemi T, Kuusio H, Sinervo T, et al: Job attitudes and well- Structural Equation Modeling 6:1–55, 1999
being among public vs private physicians: organizational justice 52. Lewis-Beck M: R-Squared; in The SAGE Encyclopedia of Social
and job control as mediators. European Journal of Public Health Science Research Methods. Edited by Lewis-Beck M, Bryman A,
21:520–525, 2011 Liao T. Thousand Oaks, CA, Sage, 2004
35. Elovainio M, Kivimäki M, Vahtera J, et al: Sleeping problems and 53. Kaplan D: Structural Equation Modeling: Foundations and Ex-
health behaviors as mediators between organizational justice and tensions. Los Angeles, Sage, 2009
health. Health Psychology 22:287–293, 2003 54. Byrne BM: Structural Equation Modeling with Mplus: Basic
36. Anderson N, West M: Measuring climate for work group in- Concepts, Applications, and Programming. New York, Routledge,
novation: development and validation of the Team Climate 2012
Inventory. Journal of Occupational Behaviour 19:235–258, 1998 55. Karasek R, Choi B, Ostergren PO, et al: Testing two methods to
37. Kivimaki M, Elovainio M: A short version of the Team Climate create comparable scale scores between the Job Content Ques-
Inventory: development and psychometric properties. Journal of tionnaire (JCQ) and JCQ-like questionnaires in the European
Occupational and Organizational Psychology 72:241–246, 1999 JACE Study. International Journal of Behavioral Medicine 14:
38. West MA: The social psychology of innovation in groups; in In- 189–201, 2007
novation and Creativity at Work: Psychological and Organizational 56. Hvidhjelm J, Sestoft D, Bjørner JB: The Aggression Observation
Strategies. Edited by West MA, Farr JL. Chichester, Wiley, 1990 Short Form identified episodes not reported on the Staff Obser-
39. Virtanen M, Kurvinen T, Terho K, et al: Work hours, work stress, vation Aggression Scale–Revised. Issues in Mental Health Nursing
and collaboration among ward staff in relation to risk of hospital- 35:464–469, 2014
associated infection among patients. Medical Care 47:310–318, 2009 57. Tenneij NH, Goedhard LE, Stolker JJ, et al: The correspondence
40. Heponiemi T, Elovainio M, Kouvonen A, et al: Ownership type and between the Staff Observation Aggression Scale–Revised and two
team climate in elderly care facilities: the moderating effect of other indicators for aggressive incidents. Archives of Psychiatric
stress factors. Journal of Advanced Nursing 68:647–657, 2012 Nursing 23:283–288, 2009
41. Goldberg D: Detecting Psychiatric Illness by Questionnaire. Lon- 58. Snyder LA, Chen PY, Vacha-Haase T: The underreporting gap in
don, Oxford University Press, 1972 aggressive incidents from geriatric patients against certified nursing
42. Kunkler J, Whittick J: Stress-management groups for nurses: prac- assistants. Violence and Victims 22:367–379, 2007
tical problems and possible solutions. Journal of Advanced Nursing 59. Stevenson KN, Jack SM, O’Mara L, et al: Registered nurses’
16:172–176, 1991 experiences of patient violence on acute care psychiatric in-
43. Paterson B, Turnbull J, Aitken I: An evaluation of a training course patient units: an interpretive descriptive study. BMC Nursing
in the short-term management of violence. Nurse Education Today 14:35, 2015
12:368–375, 1992 60. Benson A, Secker J, Balfe E, et al: Discourses of blame: accounting
44. Carson J, Cavagin J, Bunclark J, et al: Effective communication for aggression and violence on an acute mental health inpatient
in mental health nurses: did social support save the psychiatric unit. Social Science and Medicine 57:917–926, 2003
nurse? Nursing Times Research 4:31–42, 1999 61. Iennaco JD, Dixon J, Whittemore R, et al: Measurement and
45. van Weert JC, van Dulmen AM, Spreeuwenberg PM, et al: The monitoring of health care worker aggression exposure. Online
effects of the implementation of snoezelen on the quality of Journal of Issues in Nursing 18:3, 2013
working life in psychogeriatric care. International Psychogeriatrics 62. Ervasti J, Kivimäki M, Pentti J, et al: Work-related violence, life-
17:407–427, 2005 style, and health among special education teachers working in
46. Elovainio M, Salo P, Jokela M, et al: Psychosocial factors and well- Finnish basic education. Journal of School Health 82:336–343,
being among Finnish GPs and specialists: a 10-year follow-up. 2012
Occupational and Environmental Medicine 70:246–251, 2013 63. Chen WC, Hwu HG, Kung SM, et al: Prevalence and determinants
47. Holi MM, Marttunen M, Aalberg V: Comparison of the GHQ-36, the of workplace violence of health care workers in a psychiatric
GHQ-12, and the SCL-90 as psychiatric screening instruments in the hospital in Taiwan. Journal of Occupational Health 50:288–293,
Finnish population. Nordic Journal of Psychiatry 57:233–238, 2003 2008

496 ps.psychiatryonline.org Psychiatric Services 68:5, May 2017