Workplace violence targeting health care workers has been a widely recognized problem for over a decade. Results of the Bureau of Justice Statistics National Crime Victimization Survey indicated that from 2005 to 2009 the annual rate of violent victimization for nurses and physicians was 8.1 and 10.1, respectively, per 1,000 workers compared with 5.1 for all occupations (Harrell, 2011). In 2002, the National Institute for Occupational Safety and Health issued a document titled Violence: Occupational Hazards in Hospitals in an effort to increase awareness of risk factors for violence in hospital settings. In 2006, the International Council of Nurses published a position statement condemning all acts of violence against nurses. Despite an awareness of the problem, violence continues to occur. According to the most recent Bureau of Labor Statistics data, in 2014 the incidence of injury involving days away from work due to violence was 14.4 per 10,000 full-time health care and social assistance workers compared with 6.8 for workers in all industries (Bureau of Labor Statistics, 2015).
Emergency departments (EDs) have been identified as environments in which the incidence of violence is particularly high (Gerberich et al., 2005). Crilly, Chaboyer, and Creedy (2004) reported that 79% of ED nurses experienced workplace violence during a 5-month period. In other studies, 81% of nurses experienced verbal abuse (Ryan & Maguire, 2006) and 67% reported physical abuse (Gates, Ross, & McQueen, 2006). According to the Emergency Nurses Association's (ENA's) Violence Surveillance Study, 54.5% of ED nurses reported exposure to physical or verbal abuse within the previous 7 days (ENA, 2011). Violent victimization in EDs is not isolated to nurses, however. In a study of members of the Michigan College of Emergency Physicians, 74.9% of doctors reported experiencing verbal abuse and 28.1% reported physical abuse in the previous year (Kowalenko, Walters, Khare, & Compton, 2005).
Maintaining a safe work environment is a legal and ethical responsibility of health care administrators and nursing leaders. A major challenge in combating workplace violence in health care settings is the lack of documentation due to underreporting. Health care workers do not report instances of workplace violence for a variety of reasons: fear of retaliation, lack of physical injury, inconvenience, consider exposure to violence as part of the job, fear it will affect customer satisfaction scores (ENA, 2008), work experience, and manager attitude (Sato, Wakabayashi, Kiyoshi-Teo, & Fukahori, 2013). Despite the well-documented prevalence of violence in health care by the Bureaus of Labor and Justice, violent incidents are frequently not reported by nurses (Ferns, 2005); the ENA Violence Surveillance study revealed that nurses reported only 13.9% of verbal abuse and 34.4% of physical abuse to which they were exposed (ENA, 2011).
The ENA has identified the following risk factors for violence in EDs: patient access to firearms and substance use, working directly with potentially dangerous people, uncontrolled movement of the public, poor security, delays in service, crowding and uncomfortable surroundings, lack of staff training, inadequate staffing, and transporting patients (ENA, 2008). Although the majority of studies addressing interventions to reduce violence in EDs have emphasized staff training to manage aggression, the effectiveness of staff education as a solution to this problem is largely unconvincing (Anderson, FitzGerald, & Luck, 2010).
One interesting finding that has come out of the intervention literature involves the relationship between staff member perceptions of safety and actual risk. Blando, O'Hagan, Casteel, Nocera, and Peek-Asa (2013) found that workplace practices promoting staff member perceptions of safety were not correlated with lower rates of violence. Specifically, adequate security guard response time and adequate security equipment were significantly associated with staff member perception of safety, yet neither of these significantly reduced the likelihood of assault (Blando et al.).
PURPOSE
This study sought to investigate the relationships among exposure to violence, reporting of violence, tolerance to and expectation of violence, perceptions of safety, and perceived viability of interventions to reduce violence. Emergency departments are work environments in which many health care workers interact with the public and rely upon one another in order to accomplish tasks in very short periods of time. This study's inclusion of nonproviders and health care providers from various disciplines adds a multidisciplinary dimension to this phenomenon that is currently lacking.
The specific aims of this study were as follows: (a) describe the prevalence of violence experienced by ED staff members, (b) describe the perceptions of safety, tolerance to violence, and expectation of violence among ED staff members, (c) describe reporting behaviors and perceived barriers to reporting violence among ED staff members, (d) examine relationships between demographic variables, experiences of violence, tolerance to violence, perceptions of safety and reporting behaviors, and (e) determine ED staff member perceptions of viable interventions to improve workplace safety.
METHODS
Design and Setting
A cross-sectional design was used to survey all full- and part-time ED staff members in a suburban Level 1 Shock Trauma center with 48,000 annual admissions. Violence prevention strategies already existing in the ED in which this study was conducted include continuous dedicated security presence, metal detectors including wands used for ambulance arrivals, controlled access, panic buttons, state law making assault against a health care provider a felony, mandatory reporting policy, immediate triage to examination room resulting in minimal wait times, staffing levels above the national average, a locked and segregated psychiatric ED, individual patient rooms, name badges for visitors, and exclusive hiring of experienced nurses.
Upon receipt of the institutional review board approval, all ED staff members were sent an invitation to participate via organizational e-mails. The invitation to participate included a link to the confidential survey. Consent to participate was implied with completion of the voluntary survey. Instrumentation was composed of a researcher-designed survey, including demographic questions and questions regarding exposure to workplace violence based on the ENA definition, "...any physical assault; emotional or verbal abuse; or threatening, harassing or coercive behavior in the work setting that causes physical and/or emotional harm" (ENA, 2011, p. 9). See Table 1 for survey questions and response items, no pilot testing occurred. An audit of ED incident reports during the study timeframe was also performed.
Participants
All full- and part-time ED staff members from any discipline were eligible to participate. One of the researchers works at the participating hospital; the other researcher is not affiliated with the agency and had no relationship with the participants. The survey was sent to 235 eligible staff members. A total of 147 people (92 females; 55 males) completed the online survey, yielding a response rate of 63%. The majority of respondents were female (63%) and under age 40 years (55%). While most respondents reported working a day shift (n = 55, 37%), all shifts were well represented: evening (n = 32; 22%), night (n = 34; 23%), and variable or rotating shifts (n = 24; 16%). The sample included respondents from the following disciplines: registered nurse (RN) (n = 52; 35%), psychiatric RNs (n = 5; 3%), doctor of medicine or osteopathic medicine (MD/DO) (n = 28; 19%), radiology technician (n = 23; 16%), laboratory technician (n = 9, 6%), physician assistant (PA) (n = 9; 6%), unit secretary (n = 5; 3%), critical care technician (n = 4; 3%), psychiatric assessor (n = 3; 2%), and registration clerk (n = 2; 1%). Among the RNs in the sample, most had greater than 11 years of nursing experience (n = 33; 58%), but less than 10 years of experience as an ED nurse (n = 34; 60%). The majority of nurses were at least BSN prepared (n = 36; 63%) and certified emergency nurses (n = 31; 54%).
Data Analysis
Data management and statistical analysis were conducted using SPSS v 24 software (IBM, Armonk, New York). Statistical significance was set at p <= .05. Because of small representation within some disciplines, roles were collapsed as follows for statistical analysis: RNs (except psychiatric RNs) were kept as "RN"; psychiatric RNs and psychiatric assessors were combined to "psychiatric health care workers"; MDs, DOs, and PAs were combined to "providers"; critical care technicians, radiology technicians, and laboratory technicians were combined to "ancillary staff"; and registration clerks and unit secretaries were combined to "secretarial staff." Age categories were also collapsed for statistical analysis; the age groups 20-30 (n = 22; 15%), 31-40 (n = 59; 40%), and 41-50 (n = 43; 29%) years were kept as individual groups; however, age groups 51-60 (n = 18; 12%) and 61-70 (n = 5; 3%) years were combined as one group for some analyses.
Chi-square test for Independence and Fischer's Exact Test (FET) were conducted to analyze categorical data. Chi-square analyses that had more than 20% of cells with expected counts less than 5 were analyzed by FET. In addition, adjusted residuals (ARs) were examined to determine categories with significant differences between observed and expected counts. Values less than -2 or more than +2 were used as cutoffs to determine significant associations for individual cells (Sharpe, 2015).
RESULTS
Prevalence of Violence
Eighty-eight percent of respondents (n = 129) reported exposure to violence from any source within the past 6 months. All forms of violence were perpetrated by patients more frequently than by family members or visitors. Verbal abuse from a patient was the most frequent form of violence experienced, and all forms of violence perpetrated by patients were experienced to some degree by the participants in the study (see Table 2). Experiences of violence from family member/visitors were much less frequent than those by patients. No significant associations were found between exposure to any type of violence from patients or family members/visitors and age of the employee. Females were more likely to report exposure to patient sexual innuendo ([chi]2(1) = 4.16, p = .047, AR +2.0), whereas male employees were more likely to report exposure to patient physical abuse ([chi]2(1) = 6.12, p = .020, AR +2.5), patient general threats ([chi]2 (1) = 10.74, p = .002, AR +3.3), threats of lawsuit from patients ([chi]2 (1) = 4.70, p = .039, AR +2.2), and threats of lawsuits from family members/visitors ([chi]2 (1) = 5.62, p = .023, AR +2.4).
To examine the experiences of violence on various shifts, employees who reported working "variable" or "rotating' shifts were removed from the analysis to gain a clearer picture of the events occurring during each shift (subsample n = 121). As outlined in Table 2, there were no significant relationships identified between shift worked and violence perpetrated by family members/visitors. There were, however, significant relationships between shift worked and violence perpetrated by patients. Working the night shift was significantly associated with experiencing higher levels of name calling ([chi]2 (2) = 7.16, p = .029, AR +2.7), physical abuse ([chi]2 (2) = 7.73, p = .019, AR +2.8), general threats ([chi]2 (2) = 13.45, p = .001, AR +2.4), sexual innuendos ([chi]2 (2) = 6.50, p = .040, AR +2.3), and threats of litigation ([chi]2 (2) = 18.17, p < .001, AR +3.4) than working other shifts. In addition, working the day shift was significantly associated with lower levels of verbal abuse (although the chi-square test was not significant, the AR for day shift was significantly lower than expected, [chi]2(2) = 5.75, p = .056, AR -2.4), general threats, sexual innuendos, and threats of litigation than expected.
Exposure of violence by discipline is presented in Table 3. Psychiatric health care workers were the only group in which all members were exposed to some type of violence in the past 6 months. In general, ancillary health care workers and secretarial staff experienced less violence of all types than other groups. Twenty-nine percent of ancillary health care workers did not experience any violence, which was statistically higher than other groups (FET p = .003, AR +3.7). In addition, RNs, psychiatric workers, and providers experienced more violence than ancillary and secretarial staff from both patients and family members/visitors. Overall, there was a statistically significant association in exposure to the following types of violence by discipline: verbal abuse ([chi]2 (4) = 19.64, p = .001), name calling ([chi]2 (4) = 15.33, p = .003), general threats (FET p < .001), threats of lawsuit (FET p < .001), spitting (FET p = .001), and physical violence ([chi]2 (4) = 14.89, p = .005). See Table 3 for significant ARs. Of interest, only psychiatric workers experienced statistically higher frequency of exposure than predicted to threats of lawsuit ([chi]2(4) = 26.90, p < .001, AR +3.0), general threats ([chi]2(4) = 17.56, p = .002, AR +2.9), and physical violence (FET p = .006, AR +4.0), from family members and visitors.
Registered nurses (n = 57) participating in the study (including psychiatric RNs) experienced all types of violence from patients, but not from their family/visitors (see Table 4). Of note, RNs with greater than 20 years of overall (FET p = .034, AR -3.0) and ED experience (FET p = .001, AR -5.4) reported less verbal abuse than the other groups. In addition, RNs who had practiced in the ED between 5 and 10 years reported greater levels of name calling than would be expected (AR +2.5) and ED RNs with greater than 20 years of experience reported less (FET p = .014, AR -2.5). Finally, RNs who had practiced in the ED for greater than 20 years reported significantly less overall exposure to violence than nurses who had practiced in the ED for shorter amounts of time (FET p = .009, AR +4.2)
Perceptions of Safety, Expectations of, and Tolerance to Violence
The vast majority of respondents (n = 144; 98%) reported feeling safe at work. Females were more likely than males to report feeling safe at work (FET, p = .051). No differences in perception of safety were found with respect to age (FET p = .858), shift worked (FET p = 1.00), RN years of experience (FET p = 1.00), or discipline ([chi]2(4) = 5.49, p = .263). Only two respondents (1%) reported ever having missed work at this hospital because of violence from patients, visitors, or family.
Most respondents (n = 94; 64%) agreed with the statement "verbal and physical violence is an expected part of the job." No differences in expectation of violence were found based on age ([chi]2(3) = 2.57, p = .462), gender ([chi]2(1) = 2.94, p = .110), RN years of experience (FET p = .541), shift worked ([chi]2(2) = 2.45, p = .297), or discipline ([chi]2(4) = 4.95, p = .299).
Nearly 70% (n = 103) of respondents perceived their tolerance to violence as "about the same as" their coworkers, while 18% (n = 26) perceived their tolerance to violence as higher than their coworkers and 12% (n = 18) perceived their tolerance to violence as lower than their coworkers. No differences in tolerance to violence existed based on age (FET p = .141), gender ([chi]2(2) = 2.15, p = .342), RN years of experience (FET p = .211), shift worked (FET p = .910), or discipline (FET p = .651). Tolerance to violence and expectation of violence were significantly associated ([chi]2(2) = 6.19, p = .045). Respondents who agreed that violence was an expected part of the job had a higher than expected level of tolerance to violence (AR +2.0). No significant association was found between perception of safety and tolerance to violence (FET p = .213) or expectations of violence as part of the job (FET p = 1.00).
Reporting Behaviors and Perceived Barriers to Reporting
Of those who experienced violence in the past 6 months, only 3% (n = 5) of respondents formally reported all incidents, 25% (n = 37) reported some of them, and the majority (53%; n = 78) did not report any of them. Twenty-seven respondents (18%) chose "not applicable" when asked if they reported incidents of violence. However, when asked for reasons why they did not report incidences of violence, 11 respondents stated they reported all incidents, which conflicts with the direct question earlier in the survey. The most frequent reason cited for not reporting violent events was "nobody was hurt"; no respondents cited "would be perceived as weak by peers" as a reason for not reporting. Respondents were also allowed to provide "other" reasons for not reporting in an open text box. Of these responses, "not needed-normal patient behavior" (n = 12) and "lack of knowledge" (n = 10) emerged as additional reasons for not reporting. See Table 5 for a detailed list of reasons cited for not reporting events.
No significant differences in reporting behavior based on age (FET p = .666), gender (FET p = .795), shift worked (FET p = .168), or RN years of experience (FET p = .118) were found. Reporting behavior was significantly associated with discipline ([chi]2(12) = 34.81, p = .001); ancillary health care workers reported significantly less than would be expected (AR = -3.9). Reporting behaviors were not associated with health care worker's self-perception of tolerance to violence (FET p = .379) nor perceived safety at work (FET p = .063). Finally, there was a significant association between expecting violence as part of the job and reporting violent incidents (FET, p < .001); health care workers who agreed that violence was an "expected part of the job" reported significantly less violent incidents than would have been expected (AR +2.5).
An audit of formal incident reports from the ED during the same period yielded only 10 reports of violence directed toward staff members. Patients were perpetrators in all 10 of the reported cases. An audit of incident reports for the entire previous calendar year yielded only 16 total reports related to violence directed toward ED staff members.
Perceptions of Viable Interventions to Improve Safety
Fifty-four percent of the respondents (80) were aware of the Colorado state law protecting health care workers who are victims of assault. Responses to "which of the following would make our work environment safer" were zero tolerance policy (n = 80; 54%), de-escalation training (n = 69; 47%), a lighting system in front of patient rooms indicating high-risk patients (n = 24; 16%), bright-colored name tags for family members/visitors (n = 22; 15%), and different colored scrubs for high-risk patients (n = 17; 12%), and 24 respondents (16%) indicated that none of the provided options would make the work environment safer.
Participants were given the opportunity to provide suggestions to improve workplace safety via free text response; 100 participants responded. The majority (n = 60) of respondents indicated "none," "NA," "unsure," nothing was needed, the environment was safe, and/or that adequate staff members were available and prepared to handle situations. The remaining responses revolved around the following themes: security staff (n = 20), staff training (n = 20), zero tolerance policy (n = 14), visitors (n = 14), communication (n = 9), organizational culture (n = 8), police (n = 6), and situational awareness (n = 6). See Table 6 for examples of responses associated with these eight themes.
Comments related to security emphasized a need for training, for more thorough searches to be performed, and for more security officers particularly for 1:1 assignments with high-risk patients. De-escalation techniques and recognition of high-risk patients were identified as staff training needs. Remarks related to zero tolerance ranged from restricting care to offenders to the impossibility of preventing violence. Recommendations related to visitors included limiting the number of visitors and creating signage articulating behavioral expectations and consequences for breaches of those expectations while in the hospital. Comments regarding communication indicated a need for staff members with knowledge of a patient's violent behavior, or potential to become violent, to communicate that knowledge to other employees, particularly ancillary staff members and employees from other departments. In addition, there were recommendations to have a clearly articulated plan in place should a violent event occur.
Recommendations related to culture ranged from the very general need to "change the culture" to acknowledgment that "we work with high risk patients on a regular basis that can be prone to violence.... When people are altered or out of control violence can be inevitable." The importance of teamwork and a supportive staff in creating a safe work environment was evident. There were also very disparate comments related to verbal abuse, "I feel there is an idea that it is ok for staff to be verbally attacked for various reasons" and "You cannot stop or prevent people from being verbally abusive.... We should not be concerned about verbal abuse. It does not hurt us and we cannot prevent patients from speaking." These incongruent comments are reflected in other comments emphasizing a need for group consensus and cohesion regarding how to address violence within this culture. There were comments indicating a greater need for police department support to file reports against perpetrators and remove offenders from the premises. Last, there were a few comments regarding the need for staff members to recognize their role in patient escalation and improving staff member ability to identify violent individuals earlier.
DISCUSSION
Emergency department staff members in this sample expected and experienced violence at work. Violence-prevention strategies promoted in existing literature point to environmental, policy, and individual solutions. A recent study incorporating all of these elements in multiple EDs found that the comprehensive intervention had no significant effect on the occurrence of workplace violence (Gillespie, Gates, Kowalenko, Bresler, & Succop, 2014). Many recommended violence-prevention strategies already exist in the ED in which this study was conducted. Despite the presence of these prevention strategies, there was widespread exposure to workplace violence among clinicians and nonclinicians.
Prevalence of exposure to workplace violence in this study was actually slightly higher than previously published results. In addition, the nearly unanimous perception of safety in the present sample does not parallel findings from other studies. A National Emergency Department Safety Study reported that 73% of respondents felt safe always or most of the time (Kansagra et al., 2008) and likewise, only 42.3% of nurses in the ENA Violence Surveillance Study reported that they felt safe while at work (ENA, 2011). It is interesting that in this study, exposure to violence was slightly higher than previous findings, yet perception of safety was markedly higher. This reinforces existing literature indicating that staff member perception of safety and actual safety are not necessarily congruent. This finding also raises some interesting questions as to the source of this sense of safety, particularly because expectation of violence and tolerance to violence were not associated with perception of safety.
It is well established that patients are the primary source of violence directed toward health care providers. In the ED, in particular, patient violence has been associated with drug or alcohol intoxication and mental health crises (Crilly et al., 2004; Gacki-Smith et al., 2009; Gillespie, Gates, & Berry, 2013). While these might also be factors associated with visitor violence, high stress and poor communication also contribute to violent behavior in family members and visitors (Angland, Dowling, & Casey, 2014; Gillespie et al., 2013). Despite the pervasive sense of safety at work, a slight majority of respondents reported that incorporating a zero tolerance policy would make the work environment safer. What a strictly enforced zero tolerance policy would look like, however, is unclear, especially given that this facility has such a policy.
The American Nurses Association (2015) position statement on workplace violence endorses a zero tolerance policy stating, "The nursing profession will not tolerate violence of any kind from any source" (para 2). It is unethical for EDs to refuse treatment to people needing medical attention; patients in need cannot be turned away even if their behavior is out of control. In cases such as these, zero tolerance is not an option. According to the Merriam-Webster Dictionary (n.d.), tolerance is defined as "1. Willingness to accept feelings, habits, or beliefs that are different from your own, 2. The ability to accept, experience, or survive something harmful or unpleasant, 3. Your body's ability to become adjusted to something (such as a drug) so that its effects are experienced less strongly." Health care workers do in fact tolerate out-of-control behavior, particularly in the context of providing emergent medical care. Therefore, zero tolerance policies might not have a great impact on incidents of violence perpetrated by patients, but could affect the less common violence perpetrated by family members or visitors. Visitors, unlike patients, can be removed from hospital property. As is illustrated in the narrative findings of this study, cooperation from local police is necessary to enact zero tolerance policies, that is, forcibly remove disruptive visitors, and assist staff members in pressing charges against assailants. Even in situations where police departments are cooperative, zero tolerance policies have not been shown to actually prevent violence from happening (Holmes, 2006). In fact, these policies do not typically even come into effect until after intolerable behavior has occurred; intolerable behavior occurs and then offenders are dealt with per policy. This study illustrates the importance of discussing whether or not tolerance is a harmful characteristic to possess, and to clarify whether zero tolerance or zero violence is the goal in health care settings.
With respect to reporting workplace violence, there was an inconsistency noted between the number of respondents stating they reported an act of violence and the number of formal incident reports completed. This inconsistency may be indicative of the confusion that exists for the term "report." Some of the free text survey responses help clarify this discrepancy. Several respondents indicated that they reported/noted/recorded the violent act in the patient's medical chart. There is a lack of awareness that initiating a formal incident report is what is meant by reporting the incident. Many health care workers might assume that making a note in a patient chart is sufficient without realizing that data about the prevalence or incidence of violence in the workplace are not easily extracted in this manner.
According to the ENA (2008), the top barrier to reporting incidences of violence, as indicated by 45% of their sample, was fear of retaliation. Fear of retaliation and lack of support from leadership were among the least reported reasons for not reporting among this study's sample; combined these only accounted for 7% of the identified reasons for not reporting. It is possible that the observed high staff member perception of safety is linked to strong leadership support and lack of fear. Perhaps a work environment that is perceived as emotionally supportive and nonpunitive promotes a general sense of safety among staff members.
Another barrier to reporting in this particular population might actually stem from how workplace violence is defined. Many professional organizations define workplace violence as exposure to behaviors that cause emotional or physical harm. ED staff members are frequently exposed to patient and visitor behaviors that have this potential. If exposure to this type of behavior is both expected and tolerated and does not result in perceived harm, it is logical that it would not be reported. In fact, in this sample, "nobody was hurt" and "it is an expected part of the job" were the top two reasons for not reporting. By definition, if nobody is harmed, an event might not meet an individual's threshold for workplace violence which requires reporting. Some might argue that the potential for harm exists, regardless of whether or not actual harm occurred, which should therefore necessitate reporting; others might retort that the time-consuming and inconvenient nature of some reporting systems would make reporting every act in which potential harm existed prohibitive in busy EDs.
What exists are discrepancies between what organizations want staff members to report and what staff members actually report. One basic reason for this discrepancy is that organizational definitions, which drive policies, do not allow room for staff members to interpret and define events based on context. Nurses assign different meanings, and therefore respond differently, to acts of violence based on mitigating circumstances such as presenting problem, organic, and psychiatric factors (Blando et al., 2013; Luck, Jackson, & Usher, 2007). A patient who has delirium and is attempting to hit a nurse will likely be interpreted very differently than a visitor who is attempting to do the same thing. This interpretation is potentially a stronger predictor of reporting behavior than a policy mandate.
Limitations
This study's generalizability is limited by the small sample size and inclusion of only one facility. A larger sample size and inclusion of staff who worked in other EDs would have provided a broader view of workplace violence. In addition, those who opted to respond to the survey may have been different from those who did not in unforeseeable ways. With respect to the nurses in the sample, the majority of respondents were at least BSN prepared, certified in their specialty area, and had more than 11 years of experience. It is possible that nurses with less education and/or experience might differ in their exposure to workplace violence and other study variables. In addition, workplace violence can be categorized into four types based on the nature of the victim/perpetrator relationship (Lipscomb & Ghaziri, 2013). The present study addressed only types 1 (stranger/visitor) and 2 (client/customer); Type 3 workplace violence occurs between employees and Type 4 involves domestic issues that are brought into the workplace.
Implications
This study sought to pull together information regarding exposure to violence, tolerance to and expectation of violence, reporting behaviors, and perceptions of safety among a multidisciplinary sample of ED staff members. Even though the study was limited by sample size and location, specific conclusions regarding these issues among this sample of ED staff members include the following: (1) exposure to violence is not limited to direct care providers, (2) they are exposed to violence frequently, even when preventive measures are in place, (3) they expect to be exposed to violence, (4) their tolerance to violence is comparable to their peers, (5) they perceive themselves as safe, even though they are exposed to violence, and (6) they do not often report violence, primarily because nobody is hurt.
This study indicates that ancillary and secretarial ED staff members, in addition to RNs and physicians, at one Level-1 Trauma Center, were frequently exposed to violence. Therefore, discussions of violence and safety in ED settings need to include a multidisciplinary perspective. For a variety of reasons, violence in health care is not likely to be completely eliminated and therefore heightened awareness and the development of supportive work environments have been advocated (Gates, 2004). Health care workers continue to come to work despite being exposed to violent behavior and should be supported in that endeavor. More investigative energy needs to be devoted to the roles of supportive work environments and strong leadership in managing violence, particularly staff members' responses to, and mitigation of, the effects of violence when it does occur.
People respond to violence based on their interpretation of the behavior and the context in which it occurs. Most organizations, on the other hand, expect standardized responses to what they define as violence. As mentioned, if nurses equate violence with harm and no harm occurs, their responses might not involve reporting incidents. A discussion about whether or not it is necessary for nurses to report, as violence, incidents that they do not perceive as causing harm is recommended. The perception that harm did not occur also underscores the potential psychological harm that could go unrecognized and therefore unaddressed. Exposure to workplace violence has been associated with anger, posttraumatic stress symptoms (ENA, 2011; Gates, Gillespie, & Succop, 2011), and changes in overall quality of life (Wu et al., 2014). If health care workers who are exposed to violence do not associate psychological changes to their violence exposure, they may be less inclined to seek assistance.
These findings taken as a whole-staff members are exposed to violence, expect it, tolerate it, feel safe, are not harmed, and don't report it-suggest a need for leaders to support staff members in the challenging, everyday work that they do. In addition, leaders must ensure that mechanisms are in place to respond when violence that is unexpected or intolerable occurs, when staff members feel unsafe and most importantly, when they are harmed. Incidents of violence under these circumstances will be interpreted differently and ought to be responded to accordingly. The support or resources needed in these circumstances need to be assessed and put into place proactively. Resources to combat stress, fear, avoidance, anger, and other consequences of repeated exposure to disruptive behavior need to be available and easily accessible. In retrospect, an important, additional question to have asked participants who were exposed to violence would have been, "Did you have the resources you needed when this occurred?"
Finally, it might be worthwhile to reevaluate what is meant by the term "zero tolerance." Especially among those who work with patients with delirium, dementia, traumatic brain injury, psychotic disorders, or drug/alcohol intoxication, exposure to violence is not beyond the realm of possibility; in fact, it occurs and it is tolerated, perhaps because it is understandable if not expected. To propose blanket zero tolerance policies that include violent behavior perpetrated by individuals who lack the capability to comprehend what is happening and/or control their behavior is not realistic and sets individuals and organizations up for failure. If the intent is to articulate policies that direct appropriate resources to specific situations that have been identified as having a high risk for violence to occur, focusing attention on risk assessment and resource allocation having nothing to do with tolerance may be more beneficial. If the intent is to articulate policies that specifically hold accountable those with the capacity to understand said policy, as a way to deter violence or outline consequences of violent behavior, focusing communication on behavioral expectations and culpability may provide guidance in addressing challenging circumstances.
KEY POINTS
* Members of every discipline were exposed to violence; discussions of violence and safety in health care need to be multidisciplinary.
* Workplace violence occurred even when multiple violence-prevention strategies were in place.
* Despite widespread exposure to violence, there was a nearly unanimous perception of safety while at work.
* Most participants expected violence as part of their jobs and felt that their tolerance to violence was similar to their peers. Tolerance to violence and expectation of violence were significantly associated.
* Underreporting of violence might be due, in part, to how workplace violence is defined.
Acknowledgment
The authors thank the following people for their contributions to this project: Meredith Becker, Jason Brauninger, and Beth Dunn.
REFERENCES