Violence against Healthcare Workers in a Pediatric Emergency Department

Abstract

Workplace violence (WPV) by patients and visitors against nurses and physicians is a problem in adult emergency departments (ED), but largely unrecognized and unreported in pediatric EDs. The purpose of this qualitative study was to describe the WPV that occurred in a pediatric ED and the negative effects on the workers. Data included transcribed interviews with 31 pediatric ED workers, non-participant observations, digital photographs, and archival records and analyzed using a modified constant comparative analysis method. Participants perceived that both genders and all occupational groups were at risk for experiencing verbal and physical WPV. Common perpetrator characteristics were patients receiving a psychiatric evaluation and visitors exhibiting acute anxiety. Effects were experienced by the workers, perpetrators, patient bystanders, and the healthcare employer. It is concluded that WPV is a problem in this pediatric ED and interventions need to be implemented to promote the safety of the workers and patients.

Regina had been a registered nurse for 15 years before starting her orientation in this pediatric hospital. She offered to help another nurse on her team by discharging one of the patients. A few minutes later, Regina came out of the room crying, tossed the discharge papers on the counter, and said, “I have to go.” Regina left the team before her preceptor could ask any questions.

Lily, the preceptor, saw that the discharge papers had not been signed. Lily entered the patient’s room. The patient’s mother immediately said, “Why do you think you’re better than me?!” The mother was within inches of Lily’s face yelling. Lily’s heart was beating fast and she could feel the pounding in her chest. Lily didn’t understand why the mother was yelling at her. Lily left the room. The mother followed Lily still yelling. Lily sat on a chair in the walkway across from the nurse’s station. The mother was now standing over Lily still yelling. Lily looked towards the nursing station for help. The staff at the desk did not say anything. The staff did not call security for help. The staff did not do anything, except stare. Finally, the mother quit yelling and left the emergency department without the patient’s discharge instructions.

Later in the shift, another patient was placed into the same patient room. Regina said, “I can’t go back in that room.” Lily said, “Okay,” and did the nursing assessment without Regina. Lily could feel herself hurrying through the interview and assessment. Although this was a different patient and a different mother, they reminded Lily of the experience from just a few hours before. Lily’s heart rate was again speeding and her chest pounding.

The preceding anecdote is an actual occurrence of violence in a pediatric emergency department (ED) that prompted an investigation into workplace violence. The National Institute for Occupational Safety and Health (NIOSH, 2002) defines workplace violence as any violent act directed towards someone that is at work or on duty. These violent acts range from verbal violence such as racial slurs, cursing, insults, and threats (Ergun & Karadakovan, 2005; Fernandes et al., 1999; Gacki-Smith, Juarez, Boyett, Homeyer, Robinson, & MacLean, 2009; Gates, Ross, & McQueen, 2006; Levin, Hewitt, & Misner, 1998) to physical violence such as being pushed, slapped, and having objects thrown at the worker (Crilly, Chaboyer, & Creedy, 2004; Gacki-Smith et al.; Gates et al.).

Background

Gacki-Smith et al. (2009), Gates et al. (2006), and Gerberich et al. (2005) found that nurses in adult EDs are at high risk for being victims of violence by patients. Kowalenko, Walters, Khare, and Compton (2005) found a high rate of violence against ED physicians. Gates et al. reported that 67% of nurses and 51% of physicians in their study had been physically assaulted by patients. Kowalenko et al. reported that 28% of physicians in their study had been a target for physical violence and 75% a target for verbal violence by patients. Violent acts against workers in the ED setting included verbal harassment, verbal threats, sexual harassment, physical assault, confrontations after time of patient care, and stalking (Gates et al.; Kowalenko et al.). Gacki-Smith et al. found an astonishing high rate of nurses who had been physically assaulted more than 20 times (23%, n = 811) or verbal violence more than 200 times (20%, n = 604) over a three year period.

Family members or visitors of adult ED patients are also noted to be perpetrators of workplace violence. Gates et al. (2006) surveyed ED workers including physicians and nurses about violent events from visitors over the preceding six month period. The researchers found that 84% of ED workers reported being verbal harassed by visitors, 43% reported verbal threats by visitors, and 5% reported physical violence by visitors over the preceding six months. Kowalenko et al. (2005) identified violence to emergency physicians during the preceding 12-month period of their study finding that verbal violence from visitors occurred to 28% and physical violence to 11% of the physicians. Ergün and Karadakovan (2005) learned that over the preceding fiver-year period, 65% of emergency nurses experienced verbal violence and 85% of emergency nurses experienced physical violence from visitors.

Consequences following workplace violence are experienced by the ED workers, perpetrators, and healthcare organizations. Worker consequences include physical pain and injury, absenteeism, a decreased sense of safety making changes in employment, and psychological affects (AbuAlRub, Khalifa, & Habbib, 2007; Almvik, Rasmussen, & Woods, 2006; Gates, Fitzwater, & Succop, 2003; Gacki-Smith et al., 2009; Gerberich et al., 2004; Henderson, 2003; Keely, 2002; Lee, Gerberich, Waller, Anderson, & McGovern, 1999). Perpetrator consequences include physical restraint, seclusion, avoidance by healthcare workers, and having their care expedited (Aström et al., 2004; Gates, Fitzwater, & Meyer, 1999; Zun, 2003). Consequences to healthcare organizations include lawsuits, loss of productivity by the victims of workplace violence, costs associated with absenteeism and replacing workers who have quit or changed positions due to workplace violence, and the victims of workplace violence perceiving the organization negatively (Barling, Rogers, & Kelloway, 2001; Ergün & Karadakovan, 2005; Gates et al., 1999; Gerberich et al., 2004; Hegney, Eley, Plank, Buikstra, & Parker, 2006; Hesketh et al., 2003; Kowalenko et al., 2005; Mandiracioglu & Cam, 2006; Schat & Kelloway, 2003; Sofield & Salmond, 2003; Winstanley & Whittington, 2002).

It is evident from the literature that workplace violence is a problem in an adult ED environment (Crilly et al., 2004; Gacki-Smith et al., 2009; Gates et al., 2006; Kowalenko et al., 2005). However, the acknowledgement of workplace violence in the pediatric setting as a problem is much less evident. Only three studies examining the problem in pediatric settings were found in the literature. McAneney and Shaw reported in 1994 the problem of workplace violence in the pediatric ED. Eleven years later, Gerberich et al. (2005) reported that violence still occurred, albeit less often, in general pediatric settings compared to adult settings. Gacki-Smith et al. (2009) found similar results in that violence against nurses in pediatric only EDs occurred. Frequent physical violence was reported by 9.2% (n = 10) of pediatric emergency nurses and verbal violence was reported by 11.8% (n = 13) of pediatric emergency nurses. However, there are still no data that adequately describe the specific violence that occurs in a pediatric ED and how that violence may be affecting the worker victims. Therefore, it was significant for this study on workplace violence in a pediatric ED to be conducted.

Purpose of the Study

The purpose of this qualitative study is to describe the workplace violence that occurred in a pediatric ED and the negative effects of workplace violence on pediatric ED workers. The study questions are: (1) What occurs during violent events in the pediatric ED?, (2) What person, workplace, environmental, and situational factors contribute to or prevent a violent act?, and (3) What are the effects on the workers, patients, and perpetrators of violence following violent acts against ED workers initiated by pediatric patients and/or patient visitors in an urban pediatric ED?

Methods

This qualitative study used a multiple case study approach with 31 participants in the ED of a large, urban, pediatric teaching hospital located in the Midwest United States. The multiple case study approach is one of the five major approaches to qualitative inquiry (Creswell, 2007; Denzin & Lincoln, 2005; Munhall, 2007). The case study approach is used to draw inferences and interpretations from a group of persons or cases, not just a single case (Mariano, 2001). In addition, the case study approach uses multiple sources of data to confirm or refute the data from a single source (Yin, 2003).

Sample

The population of interest was the nearly 200 physicians, nurses, and allied health professionals (patient care assistants, paramedics, respiratory therapists, and child life specialists) that interacted with patients and visitors in the pediatric ED. A maximum variation sample was used; a form of purposeful sampling that increases the likelihood that study findings will reflect multiple perspectives about a phenomenon (Creswell, 2007; Lincoln & Guba, 1985; Marshall, 1996). All participants were permanently assigned to the ED, worked a minimum of 8 hours per week, were employed in the ED for at least the previous 60 days, were at least 18-years-old, and able to speak English.

Procedure for Data Collection

Sources of data were interviews with 31 ED workers, observations as a non-participant in the work environment, digital photographs, and archival records. The hospital’s archival records were hospital policies, ED policies and guidelines, and hospital and ED educational offerings.

Interviews were the first source of data collected. The current study was qualitative in nature and the participants were limited to reporting the violent event that they perceived as the worst experience over the prior six months. A few participants reported on the same violent event, but from different perspectives. Interviews were audiotaped using a tape cassette recorder and conducted individually with each participant. Field notes were documented during and after each interview. Interview periods ranged from 20 to 90 minutes with an overall average of 40 minutes. While an interview guide was used, questions evolved during the interview process. Interviews were transcribed verbatim into Microsoft Office Word 2003 (Redmond, WA) yielding 690 pages of interview data. The first author confirmed the accuracy of each transcription to the audiotape by simultaneously reading the transcript and listening to the original audiotape multiple times.

Observations were conducted of the workers, patients, and visitors in the natural environment of the ED over a 40-hour period for specific items, interactions, and events based upon factors reported in the literature that were linked to workplace violence as well as the interview data. Specific factors that were sought during the observations were events of verbal or physical violence, patients or visitors seen in the public areas that appeared to be under the influence of drugs or alcohol, interactions of workers, patients, and visitors in public areas of the ED, adequate staffing based on the visible presence or absence of ED staff, long wait times in the lobby, security or police presence, security response after receiving a call at the security desk in the lobby, metal detectors or other forms of screening for weapons, ease of patients and visitors to access the ED treatment areas, the presence of security cameras, and waiting room distractions for both adults and children. Observations were conducted in ten 4-hour blocks during day shifts, evening shifts, night shifts, weekdays, and weekends. Field notes were generated during the direct observations.

Digital photographs were taken of physical artifacts during the observation periods. Yin (2003) defined physical artifacts as physical items that can be seen in the study environment. The purposes of the digital photographs were to depict (a) the physical environment of the ED where violent events occurred, (b) important objects identified during the observations, and (c) objects that were identified as being important to the participants during the interviews.

Archival records reviewed were 499 hospital policies, department guidelines and policies, educational offerings, and department maps for their relevance to workplace violence. No access was granted for the review of other archival records such as department meeting minutes, safety event reports, and personnel records.

Human Subjects’ Protections

The study was approved by the Institutional Review Board and a waiver of written consent was authorized. A recruitment/information letter describing the study and soliciting participation was placed in the mailboxes of all ED workers (physicians, nurses, allied health professionals). Verbal consent was granted by each participant before starting the interview. Each participant chose a pseudonym in lieu of his or her actual name for the transcript.

Data Analysis

A modified version of the constant comparative method described by Lincoln and Guba (1985) was used to analyze the interview data. Data from the 1^st interview served as pilot data to practice the interview process and use of the interview guide. Data from the pilot were excluded from data analysis leaving 31 interviews for data analysis.

Data analysis began with a review of the 2^nd interview transcript while simultaneously listening to the interview audiotape; however no data coding was conducted. The 2^nd interview transcript was read once more before data coding. On the third reading, line-by-line coding was performed to identify important units of information. Units of information were identified by highlighting the text on the printed transcript using colored highlighters. Potential categories for each unit of information were written next to the unit in the right margin. The transcript was reread and recoded to ensure that no important units of information were missed. Units of information were then evaluated and compared to the original categories to determine if they were still related. A category name was applied to grouped units of information. The primary author and one of the co-authors then met to conduct investigator triangulation. Investigator triangulation occurs during data analysis when two or more researchers who have independently coded the same transcript using the same procedures meet to derive the same study findings (Sim & Sharp, 1998). During the meeting, each transcript line was compared to determine if both researchers identified the same units of information as being important. When only one investigator identified a particular unit of information, the two researchers discussed rationale and a mutual decision was agreed upon whether or not to include the unit of information. Assignment of categories for the units of information was done simultaneously during the transcript analysis. By the end of the meeting, a coding scheme for categories and overriding themes for categories was agreed upon.

The same process for listening, reading, and coding occurred during the analysis of the 3^rd interview transcript. The coding scheme was amended based on the new transcript data. The transcript for the 2^nd interview was read again and recoded to reflect the revised coding scheme. The primary author then again met with the same co-author to discuss the units of information and categories for the 2^nd and 3^rd interviews. A revised coding scheme based on data from both cases was determined. Remaining transcripts for the 4^th through 32^nd interviews followed the same format of listening, reading, and coding. After each interview transcript was coded, prior interview transcripts were read again to identify any data that may have been missed. The coding scheme constantly evolved since both researchers remained open to new and contradictory evidence with each ensuing case. This process continued until it was determined that saturation of the data had been achieved and no further interviews were warranted or solicited.

Following analysis of the interview data, method triangulation occurred. Method triangulation is the use of multiple sources of data to verify the findings between one or more other sources (Lincoln & Guba, 1985; Yin, 2003). Field notes and archival data (policies, educational offerings, and maps) were reviewed to identify if data from the interviews could be supported or refuted.

In an effort to increase the reliability and validity of the findings, the authors incorporated rigor. The four components of rigor are credibility, transferability, dependability, and confirmability. Key avenues taken to address the components of rigor were data source triangulation (comparing data from each participant transcript to the data from all prior transcripts), method triangulation (comparing interview and non-interview data), and investigator triangulation (having two researchers come to the same findings from the data). In addition, debriefing, a rich description of the findings, and an audit trail were performed.

Findings

Thirty-one workers from the pediatric ED participated in this study (see Table 1 for demographics). No participants reported Hispanic or Latino ethnicity. The mean age of participants was 33 years old ranging from 22 to 51. The mean time worked in an ED was 5 ½ years with a range of 6 months to 22 years. The mean time worked in pediatric care was 6 ½ years with a range of 6 months to 22 years. The group of variable shift workers was largely physicians.

Table 1.

Participant demographics (N=31).

Variable	Characteristic	n (%)
Role	Emergency nurse	12 (39%)
	Physician	8 (26%)
	Allied health personnel	11 (35%)
Gender	Male	3 (10%)
	Female	28 (90%)
Race	Caucasian	29 (94%)
	Non-Caucasian	2 (6%)
Age	20–29 years old	10 (32%)
	30–39 years old	15 (48%)
	40–49 years old	5 (16%)
	50–59 years old	1 (3%)
Shift	Days	8 (26%)
	Evenings	10 (32%)
	Nights	4 (13%)
	Variable	9 (29%)

Research Question 1: What Occurs During Violent Events in the Pediatric ED?

Three themes were related to the first research question: Perpetrators of Workplace Violence, Description of Violent Events, and Non-targets Becoming or Not Becoming Involved in the Event (see Table 2 for a summary of themes).

Table 2.

Themes of workplace violence in the pediatric emergency department.

Research question	Theme
What occurs during violent events in the pediatric emergency department?	Perpetrators of workplace violence Description of violent events Non-targets becoming or not becoming involved in the event
What factors contribute to or prevent a violent act?	Participants’ interpretation of workplace violence
What are the effects of workplace violence?	Personal responses to workplace violence Effects and outcomes of workplace violence

Perpetrators of workplace violence.

Participants reported several signs that occurred prior to several of the violent events: patients were being treated for a psychiatric evaluation, patients had a history of enacting violence, family members “murmured” in an aggressive manner, and family members had increasing symptoms of anxiety or acute stress. There were 16 patients that enacted violence: seven boys and nine girls. Fourteen of the patients were being treated for a psychiatric disorder, 11 of which were teenagers; however, two participants described the same event of a male teenager throwing a walker at a security officer. Family members became violent during medical and trauma care situations. Family member perpetrators were most often female: 9 patient mothers, 2 grandmothers, and 7 fathers. The two grandmothers were likely from the same event as well; both participants described a physically violent grandmother who eventually was in a physical altercation with multiple security officers.

Description of violent events.

Reports of verbal, nonverbal, and physical violence were described by participants. Verbal violence represented 50% of the violent reports and physical violence represented the other 50% of the violence reports by participants. Nonverbal violence occurred in conjunction with both verbal and physical violence. Nonverbal events perceived by participants as threatening were family members pointing a finger at the worker, invading a worker’s personal space, and blocking a worker’s ability to leave a patient room. Verbally violent events included yelling, cursing, and threatening workers with physical harm. Verbal violence occurred most often from family members (82%) compared to patients (18%). An event that included physical violence or both physical and verbal violence was coded as a physically violent event during data analysis. Physically violent events included hitting or throwing objects at workers. Physical violence occurred most often from patients (76%) and to a lesser extent from family members (24%). While observations were conducted in the main ED lobby, one patient became physically violent while in triage.

Non-targets becoming or not becoming involved in the event.

Violence didn’t just affect the ED workers most directly involved in the violent event. Several participants reported calling for additional ED workers and/or security officers to help manage violent persons. Personnel from outside the ED were also contacted: administrators, psychiatric intake workers, social workers, nursing experts, and family relation liaisons. Participants knew someone needed help when they heard loud noises or yelling or witnessed a violent act. Examples of actions taken during a violent event included setting limits with perpetrators and assisting with leather restraint application or chemical restraint administration. For example, Janice, a physician said to one patient’s mother, “‘You will not talk to us like that.’”

A few parents and family members were described as intervening with the perpetrator to control the violent behaviors. In three situations, no help was provided to the victim. First, a physician assumed that someone else had the situation under control when he passed by a room where yelling was heard. Second, a registered nurse heard yelling and physical violence and completely avoided the situation. Third, parents of a child laughed at the patient’s physically violent and self-mutilating behavior.

Research Question 2: What Factors Contribute to or Prevent a Violent Act?

One theme was related to the second research question: Participants’ Interpretation of Workplace Violence. Findings were limited to the participants’ interpretation of the factors since this was not a correlation study.

A worker’s gender was believed to influence the occurrence of violence; women may have been targeted more often for violence, but men may have more often become directly involved in physically violent events. Paramedics and patient care assistants may have had more distressing encounters with workplace violence due to the fact that they had to sit with and spend more time adjacent to violent patients. Fiona, a paramedic, talked about her experience monitoring a restrained patient:

I think the only one that it really makes the effect on is probably the person who has to sit in the room with that patient itself, or him or herself. I don’t think, I’m not a nurse, I don’t know but I don’t [think] that it affects them nearly as much as it would the techs, because they don’t have to stay in there with them and they don’t have to be worried about, “Well, what do I do if he gets out?” And “Is he going to attack me if he gets out of these restraints?” Or, you know they [nurses] give the drugs and they leave and they come in every so often to check on the patient so, I don’t think it would affect them as much.

Decreased worker experience was identified as another factor related to the risk for workplace violence. Age, race, and body size were not consistently identified as influencing a worker’s risk for workplace violence.

Patients considered by participants to be more likely to become violent included those who were informed that they were not permitted to leave and those who were being admitted for psychiatric care. Additionally, violence occurred when patients’ personal property was removed or when parents upset patients, making them angrier.

Factors thought to be related to a family member enacting violence against an ED worker included a lack of respect for women or persons in authority or being under the influence of drugs or alcohol. Not addressing and meeting the family member’s needs for the patient’s care was also thought to increase the chance for violence. Albert, a physician, said:

… in the parents mind they want to come to the emergency department. They want to um, uh get a diagnosis, get the, you know, any necessary testing, treatments, and then get out of there. And so they come and they get uh triaged you know and so you know that’s sort of, nobody comes to the ED to be triaged. Um, they come for treatment, so there’s, there’s some time there they don’t, you know that’s part of our process so that we can sort the people correctly. That’s you know, sort of not really meaningful to them. And then they have to wait in a waiting room and then they have to come in. They have to see a resident, then they might see a fellow, then they might see an attending. Um, there’s you know, delays in getting them up to their room so there’s just you know, things that are inherently you know, not, not what we, we would like them in our own system, may contribute to their feelings of frustration about what they want to get accomplished during that visit.

Several factors related to the workplace were believed to contribute to violence occurring. Access was identified as a factor for violence for two reasons. First, perpetrators were able to gain entry into the treatment area of the ED. Second, episodes of increased anxiety occurred when multiple visitors were able to congregate at a patient’s bedside. It was believed that although security personnel were responsible for controlling access from the main waiting room, visitors still found additional ways to gain entry into the treatment areas. The back hallway entrance to the ED was seen open during day shift hours; however, a non-ED worker opened the door for the first author during the night when the door was normally locked. Participants identified additional factors related to violence: perpetrators being stared at while in the public ED areas, noxious noises, crowded waiting rooms, and long waiting times for treatment.

Violence could be unexpected, considered normal, or understandable. Participants were not always certain when violence was going to occur. Several violent events that occurred during routine patient care were least expected. For instance, one physician leaned over a patient to begin her assessment. The patient reached up, grabbed the physician’s stethoscope, and started chocking her with it. In another situation, Francis, a registered nurse told a patient she would be admitted and then the patient “… hit me, I mean it was so fast that I didn’t even like, have time to like, you know how you usually when you get punched you can kind of shield yourself with your hands and I was just like, ‘Whoa!’”

Participants believed that violence in the community was increasing and this led to a greater acceptance of violence by patients and visitors as a normal way to express anger or frustration. Most participants did not condone violence, but several participants explained that they understood why the perpetrators expressed themselves with violence. Fiona, a paramedic, reflected on reasons for the violence:

Is a child acting this way because his mom was a drug addict, his dad left him and all of a sudden now he’s in foster care and nobody wants him to begin with and he’s being beaten because of it? … I think a lot of us would understand more why they’re acting out and acting the way they are, if we would take the time to, just talk to them. Or to put, say myself in their shoes and say, yeah, I would probably act that way too if I had to spend my life [the way] they spent their life.

Research Question 3: What Are the Effects of Workplace Violence?

Two themes were related to the third research question: Personal Responses to Workplace Violence and Effects and Outcomes of Workplace Violence.

Personal responses to workplace violence.

Negative consequences were experienced by nearly every participant. Physical responses included hyperarousal and a stress response such as increased heart rate and chest pounding that occurred during and after verbal and physical violence. Psychological responses such as fear, anger, and frustration also occurred after verbal and physical violence. Additional consequences to workers were avoidance of perpetrators, intrusive thoughts, and reflective thoughts about how the violent event could have been managed differently.

Effects and outcomes of workplace violence.

Patient and family member perpetrators were at risk for being physically injured or injuring others while being forcibly restrained, receiving chemical or leather restraints, being evicted from the hospital premises, or experiencing a psychological injury. Janice, a physician, said:

… some of the paramedics seem to like to be a restraint person, but I don’t necessarily think they’ve been trained properly in how to restrain a patient. Like sometimes you’re like, “Um, can you get your arm off his trachea?” … sometimes these particular couple of people I’m thinking of, they’re always first on the scene but I’m not always sure that they have been trained on how to, properly take someone down or whatever.

One patient may have experienced psychological harm by hearing several male workers talk about physically subduing her. One patient even had to appear in court due to her physical assault on an ED worker. Some family member perpetrators were forcibly restrained or escorted off hospital property.

Consequences occurred for patients not directly involved in the violent event. A few patients had care that was altered. Examples included patients not being able to access necessary treatment rooms due to a violent event occurring in the main traffic hallway and care being delayed or interrupted when an ED worker had to attend to the violent event. Several participants described that younger patients may have been scared by the violent noises they heard coming from an adjacent room or the main corridor of the ED. When the physical violence occurred in the triage area, no one in the lobby besides for the security officer and the first author seemed to notice.

Consequences to the healthcare employer included a decrease in worker productivity and a perceived poor image of the hospital by patients and visitors. Some participants reported no effect on their work productivity; however, many reported a decreased ability to focus on their work. Several participants left their work areas to informally debrief with other workers about their violent encounter or to take a break and calm down. Multiple workers left other patients or work duties to assist in the management of violent persons. In addition, it was perceived that other patients and visitors would think negatively of the ED workers and the hospital because of the violence they saw or heard while assuming that the ED worker was the aggressor.

Discussion

The participants identified signs that warn of patients more likely to become violent such as receiving a psychiatric evaluation or exhibiting signs of acute anxiety or stress. This finding was consistent with Becker and Grilo’s (2007) finding that signs for violence from adolescent patients were suicidal ideation, lack of impulse control, and self-reported drug use. In the current study, nearly all the physically violent patients were being seen for a psychiatric evaluation. Substance use was not identified as a problem for the psychiatric patients in the current study; however, this may have been due to the younger age of the patients, participants not collecting or knowing the results of urine drug screens, and participants not asking patients about substance use.

An inability to deal with crisis situations has been identified in other studies as a reason that visitors enact violence against ED workers (Catlette, 2005; Gates et al., 2006). Likewise, several participants in the current study described situations of parents expressing their anger, frustration, or inability to control a situation with verbal and/or physical violence. Participants described a number of situations in which family members acted out violently. The violent situations in the current study were consistent with those identified by other researchers as situations that increase the likelihood for violence: disagreeing with the medical plan, denying someone a service or request, perceiving wait times as long, perceiving a healthcare worker as rude or uncaring, and grieving over the death of a child (Catlette, 2005; Committee on Pediatric Emergency Medicine, 1997; Ergün & Karadakovan, 2005; Gacki-Smith et al., 2009; Gates et al., 2006; Keely, 2002; Lin & Liu, 2005; McAneney & Shaw, 1994).

Verbal violence accounted for 50% of the reports of violence in the current study with physical violence accounting for the other 50%. Other researchers found that verbal violence occurred in greater frequency than physical violence (Crilly et al., 2004; Findorff, McGovern, Wall, & Gerberich, 2005; Gacki-Smith et al., 2009; Privitera, Weisman, Cerulli, Tu, & Groman, 2005). This difference could be explained by the current study’s design. The current study was qualitative in nature and the participants were limited to reporting the violent event perceived as the worst experience over the prior six months although three participants shared two experiences with violence. In contrast, other researchers have reported the total incidence of violent events over a given time period.

Nearly all verbal violence in the current study was enacted by visitors (82%) compared to patients (18%). These results conflicted with a study of 1,209 healthcare workers that found an equally high incidence of verbal violence from adult patients (73%) and visitors (74%) (Ayranci, Yenilmez, Balci, & Kaptanoglu, 2006). Hesketh et al. (2003) reported findings in sharp contrast to both studies. These researchers found that only 11% of visitors enacted verbal threats against ED workers while 70% of patients did. Three reasons could explain these differences. First, the events of verbal violence in the current study that occurred in conjunction with physical violence were coded as physical violence. Had all events of verbal violence been tracked, the proportion of verbal violence attributed to visitors and patients may have been different. Second, Hesketh et al. defined verbal violence as verbal threats only and did not survey for other forms of verbal violence such as cursing, name calling, and racial slurs. Third, visitors may have believed that adult patients were better able to advocate for themselves without needing a visitor to do this for them compared to pediatric patients who required a parent to make decisions and provide consent for treatment. As a result, visitors of pediatric patients may have been more protective than visitors of adult patients and may have become verbally violent when the needs of the patient as they perceive them were not met.

Physical violence accounted for 50% of the violent events reported by participants; 13 participants reported physical violence by patients (76%) and four participants reported physical violence by visitors (24%). In comparison, Gates et al. (2006) found that approximately 97% of physical violence was enacted by adult patients. While the percentages were similar, it was likely that physical violence by patients may have occurred significantly more often by patients in the adult ED setting than this pediatric ED. This may be explained by a perception of ED workers that pediatric perpetrators were not actually committing violence but reacting to a threatening situation or parents and other visitors were perceived to be more threatening than pediatric patients. As a result, when asked about their worst experience with violence, physical violence by patients would not be discussed.

Participants in the current study made judgments about why perpetrators became violent and how to respond to that violence. Luck, Jackson, and Usher (2008) interviewed nurse victims in an Australian ED. They found that the nurses were more tolerant of violence from patients when the patients could not control their behavior as a result of patients with an unintentional injury, mental disorder, infection, or hypoxia. Violence was not tolerated from patients who could have controlled their violence with better decision making: patients under the influence of drugs or alcohol or angry about wait times for care. A few participants in the current study also reported that when patients were being stared at, the violent behavior seemed to escalate further. Although nurses in Luck et al.’s study tolerated violence from patients with mental disorders or acute illnesses, participants in the current study did not believe these were acceptable reasons to enact violence.

Participants stated that it was important to identify the parents’ wants and needs during a visit to the ED believing that addressing the priority needs may prevent violence. LeBlanc and Kelloway (2002) found that not providing a particular service wanted by the patient and visitors was significantly related to an outcome of violence. In addition, violence was likely to be targeted at those individuals who were perceived as the decision makers for the patient’s care (LeBlanc & Kelloway).

The influence of drugs or alcohol was believed to be a factor related to workplace violence by visitors (Catlette, 2005; Crilly et al., 2004; DuHart, 2001; Gacki-Smith et al., 2009; Gates et al., 2006; Gerberich et al., 2004; Keely, 2002; Keough, Schlomer, Bollenberg, 2003; Kowalenko et al., 2005; Lin & Liu, 2005). Several parents who were verbally violent to the ED staff in the current study were believed to be under the influence of drugs or alcohol. However, the ED workers did not involve any social workers or security officers to protect the ED workers or patients during the ED visit or subsequent discharge when the parent drove home with the patients. One physician believed the illicit substance had “worn off” by the time of patient discharge.

While observing patients and visitors waiting for their turn to receive a triage assessment, the first author perceived that the patients and visitors were very cognizant of their turn to receive care. There were several instances where a patient was inadvertently skipped due to staff misreading the “time in department” column and a parent became aggressive or upset about the situation. Bond et al. (2007) described that wait times were a significant source of stress for patients and visitors waiting for care in an ED. Gacki-Smith et al. (2009) further identified that prolonged wait times as well as the holding and boarding of patients in the ED were a perceived predictor of violent events in the ED. It became apparent in the current study that stress resulted when patients were not called back to triage in the order in which they arrived.

Hyperarousal was experienced by many of the study participants. McCaslin et al. (2006) studied the hyperarousal response of police officers following violent events. The researchers found a significantly greater hyperarousal response when the police officer was targeted for violence compared to watching or hearing about violence. Data from the current study reflected an agreement that persons directly involved with a violent event had a greater negative response than bystanders or persons watching the event occur. Of the ten participants in the current study who described a hyperarousal response following a verbally or physically violent event, eight were directly involved in the event and two were not.

The primary consequence of workplace violence for patient perpetrators in this study was the use of physical or chemical restraints, ironically in an effort to prevent violence to self or others. Zun (2003) and Aström et al. (2004) also reported that restraints were the most common consequence of violence. In a study by Almvik et al. (2006), patient seclusion was another consequence of violence for patients; however, seclusion was only discussed in the current study as a way to remove the patient perpetrator away from a group of people staring at the patient, not as an intervention for aggressive behavior. In fact, none of the patient perpetrators in the present study was ever placed in seclusion. One possible reason for the non-use of seclusion was the lack of an appropriate safety room in which to place violent patients where the patients could not harm themselves with furniture or equipment in the room.

Ergün and Karadakovan (2005) found that less than 60% of nurses in their study believed they had a right to file charges against a perpetrator for physical violence. Furthermore, 24% of the nurses believed that they would lose their job for filing charges. DuHart (2001) and Hesketh et al. (2003) discussed that a lack of reporting is common due to workers’ beliefs that they would not have the support of administrators and police would not do anything. Only one participant in the current study filed charges against a perpetrator and she reported that the police officer who responded seemed dissuasive when taking her report. However, she did have the support of the hospital security officer, a coworker, and administration when pressing charges and later during the court proceeding. Interestingly, Gacki-Smith et al. (2009) reported that nurses with no perceived barriers to reporting violent events were significantly less likely to be a recipient of workplace violence compared to nurses with perceived barriers (15.4% versus 28.5%, p < 0.001).

Ergün and Karadakovan (2005) reported that 84% of nurses in their study believed that nurses would be less productive after experiencing verbal or physical violence. Barling et al. (2001) and Sofield and Salmond (2003) reported that workers who experienced verbal violence or had a fear of future violence demonstrated decreased productivity and increased medical errors. No participants in the current study reported a medical error. This difference may have been due to the face-to-face interviews with this researcher and an underlying concern that this self-report data would be released to others. It was also possible that sufficient safety mechanisms were in place at the study site to prevent or catch medical errors before they reached a patient. Some of the workers in the current study denied any effect to their productivity; however, a greater number of participants reported some negative effect to their productivity. Negative effects included a general decreased focus and delays or avoidance in providing care to patient perpetrators or to patients whose parents were perpetrators. Avoiding patients was also a consequence described by Sofield and Salmond.

Limitations

Findings from this study are limited in that participant interviews used only self-report data; violent events were not triangulated with safety event reports or occupational injury reports. It is possible that the true scope of physical violence and physical injuries due to violence is higher than that described in this study. Another limitation is that participants may not have been able or were not ready to talk about all aspects of their experiences with workplace violence. Additionally, participants were requested to limit the discussion to the single most stressful experience with workplace violence over the previous six months. Study participation also excluded security officers, chaplains, registration personnel, hospital unit coordinators, and clinical research coordinators who may also have experienced workplace violence. The non-participant observations were conducted in an effort to minimize these limitations by determining who enacted violence against the ED workers, how the perpetrators were managed, and what factors may lead to a patient or parent to escalate with verbal or physical violence. It should be noted that only one event of violence was witnessed during the observations. The observations did provide the observer a fuller appreciation of the context in which the participants worked and helped in interpreting the interview data. The Finally, only the primary author verified the verbatim transcriptions to the audiotapes as opposed to using someone not affiliated with the study to conduct the transcription verification.

Implications for Practice

The ED is an unpredictable work setting that exposes workers to persons under extreme stress from illness or injury. Even so, ED workers have a right to safety in the workplace including freedom from acts of workplace violence by patients and visitors. Workers need to acknowledge that workplace violence can happen at any time and from any patient or visitor; therefore, all patients and visitors should be treated universally assuming that they have the potential to act out violently. It is important that all workers know the hospital’s workplace violence policies, maintain their skills of de-escalation and violence management, be cognizant of the signs that violence is likely to occur, support persons during crisis situations, and adhere to the visitor policy (Gacki-Smith et al., 2009; Gates et al., 2006; Kowalenko et al., 2005). The needs of patients and family members should be identified during the initial interview, documented in the ED record, and addressed throughout the plan of care (Kowalenko et al., 2005). Violent events may be less frequent if the needs are clearly addressed by ED workers. Ignoring the needs may result in an event or escalation of violence.

Colleagues and supervisors who are notified that a violent event occurred need to check on the safety and well-being of the worker who was verbally or physically assaulted (Gacki-Smith et al., 2009). The workers need to be provided an opportunity to informally debrief about their experience of violence and be allowed to leave the patient care area to take a break. The victim’s patient assignment should be changed such that the worker is not directly interacting again with the perpetrator during that perpetrator’s ED visit. However, this may be difficult when the victim is a physician and he or she is the only physician on staff at the time of the violent event. When this occurs, another ED worker needs to go into the room with the physician during further patient encounters. In addition, a potentially helpful measure would be to limit a worker sitting with a restrained patient for longer than thirty minutes; thereby, reducing the time period that a single worker is potentially exposed to verbal violence before being relieved for a break. Nearly 60% (n = 2,031) of the participants in a national cross sectional sample of emergency nurses related that a staffing shortage was related to workplace violence (Gacki-Smith et al., 2009). As a result, the ability to rotate staff from a violent patient’s bedside may be severely limited. Finally, every violent event needs to be documented in the hospital’s safety event reporting system and police charges pressed for events of physical violence (Peek-Asa, Cubbin, & Hubbell, 2002).

Implications for Healthcare Employers

The OSHA (2002, 2004) and NIOSH (2006) have been clear on the need to educate healthcare workers on workplace violence prevention. Healthcare employers have an obligation to ensure that all workers have readily available access to and understand workplace violence policies. Employers need to track, manage, and actively prevent verbally and physically violent events the same as they would for clinical indicators such as ventilator-acquired pneumonia and urinary tract infections. Policies need to clearly reflect an ongoing requirement for education on workplace violence prevention (Gacki-Smith et al., 2009). Components of the education program may include hospital policies, debriefing training, non-violent crisis intervention techniques, victim and bystander support, and mandatory reporting practices. The education needs to be mandatory and verified on an annual basis along with other annual mandatory trainings such as blood borne pathogen training.

There are additional implications for healthcare employers that may be effective for managing precursors to workplace violence. A dedicated volunteer or hospital employee would be beneficial for violence prevention by working with patients and visitors in the waiting room, providing distractions, meeting basic needs such as warm blankets, and serving as a liaison between the ED workers and family units. Access to treatment areas needs to be restricted. Signage to reflect visitor restrictions should be clearly visible to patients and visitors. Psychiatric intake workers can be hired to work in the ED around the clock to expedite psychiatric patient assessments, consult with aggressive patients and visitors, and provide coaching to workers involved in violent or pre-violent events. Most importantly, the rights of ED workers need to be as equally respected as the rights of patients. This can be demonstrated by the hospital formally adopting a zero tolerance policy for workplace violence that is clearly communicated to patients and visitors on arrival to the hospital and workers upon hiring (Gates et al., 2006).

There are also several implications for after a violent event has occurred. Verbal reports of physical or verbal workplace violence need to be confirmed that they have been entered into the hospital’s safety event reporting system. Follow-up assessments with workers that have been victimized are important to ensure they are able to return to work without restriction. Finally, workers subpoenaed to testify related to the violent event should be paid for their time at court.

Recommendations for Future Research

Findings from this study suggest the need for two additional research studies. First, the incidence of workplace violence in a pediatric ED needs to be determined. It is anticipated that workplace violence is a much greater problem than currently appreciated in the literature albeit of less frequency than in adult EDs. Second, there is a need to develop and test interventions aimed at reducing the incidence of workplace violence and decreasing the severity of negative consequences of workplace violence in a pediatric ED. It is believed that an intervention developed through an action research process in collaboration with ED stakeholders is likely to demonstrate a significant reduction in the incidence and severity of violent events.