Nurses have the highest rates of aggression exposure of all clinical workers, possibly due to the type and amount of patient contact. Aggression in healthcare settings is a complex problem that affects patient outcomes and worker health and requires careful consideration of measurement, monitoring, and intervention. Health care workplace aggression exposure is common; more than half of workers report past year exposure (Findorff, McGovern, Wall, & Gerberich, 2005; Winstanley & Whittington, 2004). Nurses have the highest rates of aggression exposure of all clinical workers, possibly due to the type and amount of patient contact (Bureau of Labor Statistics (BLS), 2011; Chapman, Styles, Perry, & Combs 2009; Findorff, McGovern, Wall, Gerberich, & Alexander, 2004; Gates, Ross, & McQueen, 2006; Winstanley & Whittington, 2004). Patients are the primary perpetrators of aggression toward nurses and other providers, although family and visitors may also behave aggressively (BLS, 2011; Gerberich et al., 2004; Hesketh et al., 2003; Hodgson et al., 2004; O’Connell, Young, Brooks, Hutchings, & Lofthouse, 2000; Winstanley & Whittington, 2004). Despite the high prevalence of these acts, it is difficult to obtain accurate estimates of both patient aggression and rates of health care worker aggression exposure.

In addition to health problems from physical injury, worker responses to violence may negatively impact the work environment by creating fearfulness; low morale and productivity; absenteeism; and turnover. Previous estimates of United States (U.S.) worker rates of exposure to physical aggression were approximately six percent of all employees; aggression rates in health care workers were and are higher (Schat, Frone, & Kelloway, 2006). Weekly verbal aggression exposure in U.S. workers was ten times higher than physical aggression (13% verbal versus 1.3% physical; Schat et al. 2006). This is particularly disturbing given high physical injury rates and studies suggesting that threatened violence may cause greater psychological effects on workers than physical aggression (Gerberich et al., 2004; Whittington, Shuttleworth, & Hill, 1996). In addition to health problems from physical injury, worker responses to violence may negatively impact the work environment by creating fearfulness; low morale and productivity; absenteeism; and turnover (Figure 1). For these reasons, accurate monitoring of worker aggression exposure is needed to better understand the prevalence, health effects, and organizational costs in health care settings.

Figure 1

Bureau of Labor Statistics, U.S. Department of Labor (2006a, 2006b).  [View full size figure]

This article will provide an overview of aggression exposure in health care and review the challenges of monitoring and measurement of patient aggression. Discussion of a pilot study will present insights gained from using a novel measure of aggression, handheld counters. The conclusion will discuss clinical, and research implications.

Overview of Aggression Exposure in Health Care

Consideration of aggression as a continuum of behaviors enhances measurement and intervention efforts... For the purpose of this discussion, aggression is defined broadly as “any verbal, nonverbal, or physical behavior that was threatening…or actually did harm to self, others, or property” (as cited by Morrison, 1990, p. 67; based on a 1974 American Psychiatric Association Report). The definition includes low level, minor events (e.g., yelling and threatening) and physical events (e.g., throwing things, hitting, biting, kicking, choking). Consideration of aggression as a continuum of behaviors enhances measurement and intervention efforts, which vary based on the aggression type, behavior trajectory, and clinical setting. This section will introduce some risk factors for aggression and discuss common effects of worker exposure and aggression measurement challenges.

Risk Factors for Aggression: Who, What, When, Where, Why?

Many factors increase the risk of aggressive behavior in health care settings. The Occupational Safety and Health Administration (OSHA) identifies risk factors for violence in health care including free movement of the public in health care settings; long waits for services; presence of distraught patients and family members; low staffing levels; isolated work with clients; lack of workplace aggression training; use of hospitals by the criminal justice system; and insufficient community-based care for the mentally ill (OSHA, 2004).

Perpetrators of aggression are most often patients, but may also include family members, visitors, and other workers. Who? One factor to consider is who commits acts of aggression. Perpetrators of aggression are most often patients, but may also include family members, visitors, and other workers. The variety of possibilities for violent acts can lead to more challenges to accurate measurement by population and/or setting. For example, patients in one setting may tend to be less aggressive; however, family and visitor aggression in this area may be higher (Chapman et al., 2009; Gates et al., 2006; Winstanley & Whittington, 2004).

What? Societal factors, including handgun and weapon access among patients and visitors to health care settings, may increase risk of work related assaults (OSHA, 2004).  Review of several studies revealed that more than half of clinical workers reported exposure to verbal aggression, and 7% to 52% reported physical aggression in the past year from patients, visitors, or co-workers (Table 1). While patient aggression (all types) is overall most frequent, the most common aggressive behaviors by family and visitors are verbal aggression and threatening (Chapman et al., 2009; Gates et al., 2006; Roche, Diers, Duffield, & Catling-Paull, 2010; Winstanley & Whittington, 2004).

Table 1. Hospital Workplace Aggression Exposure, By Type of Aggression and Time Period

When? Aggressive or combative behavior typically occurs when care is being delayed, delivered, or necessary care is refused (Whittington et al., 1996). Patient aggression risk related to these factors may be due to the foreign environment; equipment; discomfort with position; medication effects; intoxication or withdrawal from substances or alcohol; an existing mental health disorder; and confusion or cognitive impairment (e.g., delirium, dementia) (Hahn et al., 2008; Whittington et al., 1996). Nursing activities involved in delivering care, such as helping a patient complete hygiene, eat, dress, use the bathroom, or reorient to the environment can precipitate aggressive or combative behavior (Whittington et al., 1996). Family and visitor aggression may relate to the perception that a vulnerable patient needs protection or is not receiving needed care (Winstanley & Whittington, 2004).

Where & Why? All clinical settings involve risk of aggression exposure, with risk and reasons varying by clinical setting. For example, workers in one area may be more at risk for exposure to physical acts of aggression, and in another area, the risk is greater for exposure to verbal aggression. In one study of 113 nurses, Chapman et al. (2009) found that all nurses in medical, emergency, and psychiatric settings reported exposure to aggression; from 73% to 100% reported threats, and from 50% to 85%, physical aggression exposure. The Minnesota Nurse’s Study found risk of physical aggression exposure is two times higher in psychiatric versus medical-surgical settings, and risk of verbal aggression is nearly three times higher (Gerberich et al., 2004).

Why is risk for worker aggression exposure highest in psychiatric settings (see Figures 2 and 3)? High rates in psychiatric settings relate to both the nature of the disorders treated and the care environment. Psychiatric unit restrictions often infringe on individual freedom (e.g., locked doors, involuntary admission) to protect patient safety (Kling, Yassi, Smailes, Lovato, & Koehoorn, 2009). Unit-based processes, including unit routine and structure; level of stimulation; management of transitions, patient autonomy, and safety; and provision of basic living and care needs, can impact occurrence of aggression (Hamrin, Iennaco, & Olsen, 2009). Studies have found a negative correlation between staffing levels and aggression rates and a direct relationship between longer lengths of stay and higher rates of aggression (Lehmann, McCormick, Kizer, 1999). A recent review identified patient risk factors including impaired cognition; thought disorder or psychosis (i.e., hallucinations, delusions, paranoia); and substance intoxication or withdrawal (Hamrin et al., 2009). These patient risk factors also play a role in aggression risk in other health care settings.

Figure 2

Bureau of Labor Statistics, 2011; Harrell, 2011 [View full size figure ]

Emergency department (ED) workers have high exposure to verbal and physical aggression, where rates of verbal aggression exposure are as high as 75% to 100% and a physical aggression exposure rate of 67%, (Gates et al., 2006; Whittington et al., 1996; Winstanley & Whittington, 2004). Why such high rates? Emergency departments involve life or death crises; waiting for care (and watching others more critical receive care first); and high anxiety levels of both patients and accompanying family members or friends (OSHA, 2004; Whittington et al., 1996; Winstanley & Whittington, 2004).

Figure 3.

[View full size figure]

Other setting-based reasons for aggression exposure vary by patient population and kind of care provided. High rates of verbal aggression and threatening by family and visitors have been identified in obstetrics and pediatrics for many of the same reasons (Chapman et al., 2009; Gates et al., 2006; Winstanley & Whittington, 2004). At the same time, studies found that obstetrics and pediatrics are lower risk settings for patient aggression, with lower rates of worker exposure to patient verbal (12-33%) and physical (6-10%) aggression (Whittington et al., 1996; Winstanley & Whittington, 2004). Worker exposure rates and risk of physical aggression in long term care and geriatric settings were high when compared to medical-surgical settings and may relate to refusal of required care and impaired cognition (see Table 2). Some studies found medical unit workers had rates of verbal (39-100%) and physical aggression exposure (39 -75%) similar to ED rates (Chapman et al., 2009; Whittington et al., 1996; Winstanley & Whittington, 2004). This variability in exposure rates suggests potential unit level differences in aggression occurrence (Roche et al., 2010).

Table 2. Frequency of Worker Exposure by Unit and Type of Aggressive Event (past year)

Common Effects of Worker Exposure

Effects of aggression exposure include psychological and physical injury and often result in workplace problems. Effects of aggression exposure include psychological and physical injury and often result in workplace problems. Many psychological repercussions of aggression exposure in workers affect the workplace itself with such outcomes as decreased job satisfaction and job performance (Hesketh et al., 2003; Merecz, Drabek & Moscicka, 2009; Schat & Frone, 2011); absenteeism; and fears of returning to work (see Figure 1).

The type of exposure can impact the outcome. Gerberich et al. (2004) found that workers exposed to non-physical violence had higher rates of job transfer, quitting, and restricted work than those exposed to physical violence. Nachreiner, Gerberich, Ryan, & McGovern (2007) found those assaulted in the past year had greater work stress, lower morale, less respect for staff, and perceived less supervisory support.

In general, nurses exposed to aggression felt less safe at work compared to unexposed staff members. The number of violent exposures was inversely correlated with feelings of job safety and satisfaction (ED Management, 2007; Gates et al., 2006; Kansagra et al., 2008). Also noted were changed attitudes toward patients, psychological withdrawal, and burnout (Merecz et al., 2009).

Emotional reactions common to health care workers exposed to aggression are anger, frustration, irritability, sadness, hurt, resentment, and helplessness. Emotional reactions common to health care workers exposed to aggression are anger, frustration, irritability, sadness, hurt, resentment, and helplessness (Gerberich et al., 2004; Lanza, 1983; O’Connell et al., 2000). Anxiety, depression, and post-traumatic stress disorder (PTSD) symptoms have been found in nurses exposed to aggression (Needham, Abderhalden, Halfens, Fischer, & Dassen 2005; Richter & Berger, 2006). Whittington et al. (1996) found that nurses who had been threatened had significantly higher scores on the General Health Questionnaire (GHQ), a measure of psychological distress. Similarly, Gerberich et al. (2004) found higher rates of psychological symptoms in those exposed to non-physical aggression. Stone, McMillan, and Hazelton (2010) found nurses experienced higher distress from verbal aggression perpetrated by relatives or caregivers of patients.

Worker perceptions are likely important to understand the meaning front line workers assign to aggression exposure, and thus may influence reporting and latent psychological responses to exposure (Luck, Jackson, & Usher, 2008). With significant risk to physical and psychological health, accurately measuring aggression and aggression exposure is important. The next section presents some of the challenges associated with accurate measurement of worker exposure to aggressive behaviors.

Measuring Aggression

Challenges of Measurement

To measure acts of and thus potential exposure to aggression, rate data, or frequency of aggressive acts, are collected. The purpose of data collection to measure patient aggression and exposure to worker aggression may include research studies, quality improvement, safety monitoring, and intervention evaluation.

Challenges in measuring acts of aggression include lack of available information, use of indirect measures, and reluctance by victims to report an incident. Challenges in measuring acts of aggression include lack of available information, use of indirect measures, and reluctance by victims to report an incident. Little information is publicly available about hospital workplace aggression. Currently only national rates of crime victimization and occupational injury are publicly available to attempt to quantify workplace aggression.

Available information. Nationally monitored reports of workplace violence can be one measure of aggression. These reports include the Department of Justice rates of worker assault from the National Crime Victimization Survey (NCVS) and the Bureau of Labor Statistics frequency of lost work time due to injury or illness. Rates of occupational injury and illness due to violent acts are 2 per 10,000 workers for all private industry with higher rates (9.7 per 10,000) in health care and social assistance workers. Much higher rates occur in nursing personnel (37.7 per 10,000) and psychiatric and substance abuse hospitals (77.7 per 10,000 workers) (Figure 2). The NCVS, a community-based survey identifies that only 47% of workplace victimization is reported to the police (Harrell, 2011). The NCVS reported rates of victimization by aggression in all medical occupations of 6.5 per 1,000 employed persons between 2005 and 2009, while in mental health occupations the rate was 20.5 per 1,000 (see Figure 3).

Indirect methods and/or reluctance to report. Indirect methods to measure aggression in health care may include hospital-based reporting of restraint or seclusion (which assumes this intervention resulted from an aggressive act) and incident reports (which are proprietary information and difficult to obtain). Given the challenges of directly measuring acts of aggression, an indirect method sometimes used to quantify patient aggression is the restraint rate, which may be used as a proxy (substitute) measure for worker exposure. This method is not without concerns. First, use of restraint does not always indicate an act of aggression. Second, use of restraint rates in general hospital versus psychiatric hospital settings is problematic because the purpose of restraint has been found to differ. For example, in U.S. acute care hospital intensive care, medical, or surgical units, restraints were used to prevent disruption of treatment (74.9% of events), to manage patients with confusion (25.4% of events), and for fall prevention (17.6% of events) (Minnick, Mion, Johnson, Catrambone, & Leipzig, 2007). In psychiatric settings, restraint is used primarily to manage unsafe or aggressive behaviors toward oneself and/or others.

... both indirect and self-report methods are likely to underestimate the actual amount of exposure to aggression by health care workers. Indirect measures can only suggest acts of aggression. Direct measures are only as good as the reporting... Even this proxy may underestimate acts of aggression because in many general hospital and psychiatric settings restraint use has been reduced or eliminated (Evans, Wood, & Lambert, 2002; Hellerstein, Staub, & Lequesne, 2007; Smith et al., 2005; Steinert et al., 2010). In a study of state psychiatric settings, seclusion use was reduced by 36% and restraint use by 46% over a four year period (Schacht, 2006). Finally, restraint rates do not accurately portray the magnitude of aggressive behavior, although they do show setting-based differences in the characteristics of aggression (e.g., threats, verbal, physical).

Other research has considered workplace aggression more directly using scales and retrospective surveys. Table 1 lists several studies that illustrate the exposure to aggression in the hospital workplace by type of aggression and time period, and Table 2 shows frequency of worker exposure by unit and type of aggressive event. Studies using retrospective surveys indicated high health care worker aggression exposure. Fifty to 92% of responders reported verbal aggression exposure in the past year; 23% to 69% reported threats and from 7% to 52% reported physical aggression exposure (see Table 1). Several studies specifically looked at exposure to aggression in the week prior to data collection and found that in that week, 14% of responders were exposed to physical violence and 46% to more than one type of violence (Duncan, 2001; Hesketh et al., 2003; Roche et al., 2010).

 In sum, both indirect and self-report methods are likely to underestimate the actual amount of exposure to aggression by health care workers. Indirect measures can only suggest acts of aggression. Direct measures are only as good as the reporting, and many workers do not report aggression; they perceive it as ‘part of the job’ (Lanza, Schmidt, McMillan, Demaio, & Forester, 2011; Poster & Ryan, 1994) or the reason for patient hospitalization. A recent study found just 16% of events were reported using incident reports, and only half the time were the events verbally communicated to senior staff (Chapman et al., 2009). Only one-third of workers reported events not resulting in injury or restraint (Gates et al., 2006).

Common Aggression Measures

A benefit of using common measures... is the ability to compare rates identified with rates obtained from other settings and studies... However, these methods of recording patient aggression were not designed to record worker exposure.Common measures to quantify patient and worker aggression are listed in Table 3. A benefit of using common measures such as the Patient-Staff Conflict Checklist Shift Report (PCC-SR) and the Staff Observation Aggression Scale-Revised (SOAS-R) is the ability to compare rates identified with rates obtained from other settings and studies. For example, a review of the use of the SOAS-R is available from 15 years of research use, and offers comparative information internationally regarding aggressive behavior. However, these methods of recording patient aggression were not designed to record worker exposure.

A number of studies have developed or used scales that ask workers to identify aggression exposure using self report. Eliminating misclassification of exposure by obtaining a count of actual aggression exposure across each shift from workers would strengthen research about exposure to workplace violence. Given these barriers to an accurate count of violent and aggressive events in health care settings, new methods to objectively measure worker aggression exposure are needed.

Table 3. Common Aggression Measures

The Need for Monitoring and Improved Measurement of Aggression in Health care

Two interlocking areas must be addressed to improve understanding of patient aggression and worker aggression exposure. First, monitoring or surveillance of aggressive events in health care systems should be routine. This will positively impact efforts to prevent patient aggression and its many important consequences, including worker health and safety. Second, common measures must be developed and used to identify events, their characteristics, interventions, related exposures and patient and worker outcomes or sequelae from events.

Monitoring

Monitoring to acquire the baseline incidence and prevalence of all types of aggressive events across the entire continuum of aggression (e.g., verbal, physical) would be useful to evaluate the effectiveness of preventive intervention. For many years, researchers and clinicians have identified the need to monitor aggression in hospitals. However, monitoring is not yet comprehensive or standardized. This hinders efforts to evaluate change due to interventions and to monitor worker safety, even in psychiatric settings where rates and safety risks tend to be highest (Arnetz, Aranyos, Ager, & Upfal, 2011; Hahn et al., 2008; Lanza, 1988; Lanza, 1991; Lion, Snyder, Merrill, 1981; Nijman, Bowers, Oud, & Jansen, 2005a). The full range of aggression is not routinely monitored. Ignoring lower level aggressive events is detrimental to measurement, intervention, and improving patient and worker safety. Without routine comprehensive measurement of aggressive events, one cannot determine the probability of a particular type of aggressive event or a positive or negative event outcome. Monitoring to acquire the baseline incidence and prevalence of all types of aggressive events across the entire continuum of aggression (e.g., verbal, physical) would be useful to evaluate the effectiveness of preventive intervention.

National standards require hospitals to monitor events of injury or illness to workers and patient restraint (OSHA, 2001; The Joint Commission, 2010; 2011); however due to the challenges discussed above, these often identify only a fraction of aggressive events. They typically represent serious events, higher on the aggression continuum, which have had a negative or untoward outcome in terms of worker or patient safety and liability (Arnetz, 1998; Ross, Bowers, & Stewart, 2012). Rate differences between standard and scale measures demonstrate that the majority of events are successfully resolved short of restraint or injury and do not have negative outcomes. Thus efforts to better understand successful resolution of aggression via monitoring and evaluation may be of low priority. However, due to likely under-reporting; the possibility of multiple negative outcomes; and many events that are ‘near misses’ of sentinel events (e.g., restraint, injury), interventions to support a successful resolution to an aggressive event should be better understood. We also know that psychological effects after non-physical events are greater than after physical events; overlooking this aspect limits our understanding of how to reduce worker risk (Gerberich et al., 2004).

Measurement

...successfully resolved events are not typically examined to learn what worked and thus insight from the majority of aggressive events and the interventions that resolved them is not systematically evaluated. If an aggressive event results in seclusion and restraint, there is often debriefing of staff involved. Debriefing offers a chance for reflection on the episode, learning from the experience, and planning for future prevention and intervention. Events where interventions were less successful are carefully examined. However, successfully resolved events are not typically examined to learn what worked and thus insight from the majority of aggressive events and the interventions that resolved them is not systematically evaluated. To change this, measures need to identify patient aggression and worker exposure as well as antecedents, interventions, and outcomes for patients and workers. Identification of the trajectory of behavior and intervention(s) across all levels of aggressive events can provide insight to improve intervention. Current measures tend to either identify only patient aggression or worker exposure; without providing a comprehensive understanding.

Current measures tend to either identify only patient aggression or worker exposure; without providing a comprehensive understanding. Improvement of the measurement of aggression may help eliminate undercounting of aggressive events in health care settings. Required debriefings after sentinel events aid in understanding root causes and planning effective intervention. This process should also be studied, if possible, using an experimental design to examine other events across the aggression continuum and potential interventions. Comprehensive data collected from both the patient and worker safety perspective may improve approaches to patient aggression management and workplace safety and health.

An appropriate first goal is to use a measure that will help researchers quantify, as accurately as possible, the number and characteristics of aggressive events that health care providers are exposed to in the workplace. Counters to measure patient aggression have been introduced (Lanza, 2009; Ridenour, 2009), although reliability and validity have not been described. For these reasons, a pilot study was conducted by the authors to consider the feasibility of this relatively new exposure measure, handheld counters.

The Aggression Exposure Pilot and Lessons Learned

The Aggression Exposure Pilot Study was conducted to obtain measures of worker exposure to patient aggression using a combination of a handheld counter and a log. The purpose was to assess the usefulness and efficiency of the counters with the intent to use them in a larger participatory action research study designed to examine exposure to workplace aggression in a more comprehensive way.The study was approved by the Institutional Review Board at both the hospital and the researchers’ university.

Setting and Participants

The pilot study took place with clinical workers from the acute inpatient psychiatric setting, recruited by trained research staff during change of shift meetings. Clinical workers were provided with a brief explanation of the pilot purpose and methods. We recruited 11 workers across multiple shifts who agreed to allow the researchers to shadow them as they physically used the handheld counters to record their aggression exposure. Researchers also recorded field notes during the shadowing time.

Data Collection

A handheld counter, researcher-created log, the SOAS-R and PCC-SR scales, and unit-level rates of restraint and injury were the data collection methods piloted in this study. The small hand held counter (see author photo in Figure 4) was attached to a coil bracelet with one side marked ‘v’ for verbal events and the other side ‘p’ for physical events. As a clinical worker was exposed to an event, the researcher ‘clicked’ the counters to register the event and immediately after an event documented it on the log. Logs provided information on the patient; severity of event (range 1 to 5); date and time; risk factors present; and a brief event description. If no events occurred, a ‘0’ was entered on the log (see Figure 5). The counters provided a worker exposure rate each shift. If completed by all workers, a patient rate of aggression and a trajectory of events across shifts can then be calculated by compiling the events by time, patient demographic information, and event descriptors. The SOAS-R and PCC-SR scales were completed by the participants. Routinely documented unit rates of restraint and injury were also collected to allow comparison between all measures during the pilot data collection period.

Figure 4. Example of Handheld Counter

Pilot Results

Worker event exposure in the pilot study demonstrated great variability. For example on the first day of the pilot, the participating clinical workers and researchers were not exposed, but a large number of other workers that shift were exposed to a patient who was yelling loudly, and then disrobed and jumped on the nurses’ station. Over the prior six months there had been an average of two assaults and one injury per week due to aggression. Table 4 lists data from the pilot shifts. There were no restraints or injury reports returned over the 11 shifts. Use of counters indicated that four workers had exposure to 8 events (6 verbal, 1 both verbal and physical, 1 physical), while seven workers had no aggression exposure during the piloted shifts. Conversely, the end of shift form, the PCC-SR, indicated only 3 events, 1 verbal and 2 physical, across pilot shifts (Table 4).

Table 4. Results of Piloting of Aggression Measures

Lessons Learned

The pilot study provided insight into data collection using the handheld counters and logs and similarities and differences from other data collection methods. A review of data collected and field notes recorded during the piloting of all measures helped to identify strengths and limitations of each measure and considerations for use in other clinical settings. An asset of use of the SOAS-R was the identification of the target, consequences, and interventions related to each event. This information was not consistently available from the counter/logs nor the PCC-SR. If completed consistently, the SOAS-R should identify the same events as the counter and logs, thus this would likely result in underreporting using the SOAS-R due to redundancy.

The PCC-SR identified aggressive events and patient conflict and containment, and nursing care activities (i.e. hygiene, medication administration), providing a measure of some nursing care demands. Nursing staff participation and ‘buy-in’ was important, and accuracy was dependent on the charge nurse’s role and team communication. For example, on some units and shifts the charge nurse is visibly engaged in the milieu, while in others the charge nurse is scheduled to attend meetings and complete other duties. Despite this variability by unit, the charge nurse was aware of more severe events, providing information on the more serious or extreme events and neglecting lower level events like verbal aggression or refusal of meals or hygiene. The PCC-SR does not require the nurse to consult with all staff members to record events, relying on unit communication processes and thus may result in inaccuracies (i.e. minor or moderate incidents that are resolved may not be communicated). Single events involving many behaviors cannot be identified from this scale, thus if there were 3 verbal, 4 physical and 2 refusals of meal identified, this might represent a total of 4 or 9 individual events. The PCC-SR picked up events the other measures did not, but with the limitation that they were neither patient- nor worker-specific events.

Use of counters measured worker exposure, and provided recognition of the difficult behaviors workers manage daily. Compared to the SOAS-R and PCC-SR, the counter logs collected worker exposure and narrative information about events. Use of counters measured worker exposure, and provided recognition of the difficult behaviors workers manage daily. Some staff felt using counters acknowledged and validated the work rarely noticed by themselves or others. In addition staff often identified renewed awareness of the need for good self care given the regular exposure to difficult behavior at work. To provide a more comprehensive perspective, logs should include post event information such as consequences (e.g., injury, damage, adverse patient or staff effects); interventions; whether events were documented or reported; and debriefing or care planning.

Research and Clinical Practice Implications

Better measurement of aggressive events would contribute to greater understanding of the many aspects of this complex issue. The challenges are how to best manage or prevent aggressive patient behavior; how to reduce worker exposure to aggression; and how to assist workers to recover from and prevent ongoing health or work problems. Several research and practice implications can be suggested as a result of this pilot study.

...measures must be simple and easy for workers to use as they go about their daily work. Strategies are needed to better measure aggression exposure in hospital settings. First, measures must be simple and easy for workers to use as they go about their daily work. Retrospective data collection involving documentation on forms will have the propensity for under-reporting of rates of aggression. Methods providing within shift monitoring are useful. Any method that decreases paperwork would improve feasibility and have an advantage over existing measures. Unit-based measures (e.g., the PCC-SR) may not be as useful in general hospital settings where the charge nurse may not have broad knowledge of problem behaviors occurring.

The measures described were used in psychiatric inpatient units and were designed to be specific to this setting. One improvement would be to include more comprehensive information about events, as outlined below, in a single measure. With modifications, the measure could be tailored to needs in other clinical areas. Suggested modifications for measures include:

• Recording of event exposure of staff and determining both the number of events per unit, patient, and worker.
• Including information pertaining to the perpetrator and characteristics of the aggressive event.
• Representing risk factors, antecedents, and precipitants to events more broadly to include those common in general hospitals such as discomfort, confusion, or refusal of required care. Identification of the primary diagnosis of the patient would also provide the ability to better understand risks associated with hospitalization in various clinical settings.
• Capturing a variety of nursing activities that may relate to conflict and containment events in other settings, such as turning patients, helping patients ambulate, or performing a painful procedure. Describing these activities in event reports will increase our understanding of high risk activities.
• Identifying interventions used to manage aggressive behavior and outcomes or consequences of the event to both patient and workers.

Given the differences noted earlier in various settings, further development of setting-specific measures that incorporate any known factors would be useful.

Conclusion

Results of piloting measures suggest there is benefit to using a device such as a handheld counter combined with a paper log to record events as they occur during a shift, providing greater accuracy of exposure. Focusing on obtaining comprehensive aggressive event information will provide insight into understanding aggressive behaviors and intervention in clinical practice. The feasibility of handheld counters should be evaluated in non-psychiatric settings. Once rates of patient aggressive behaviors are better known and information on the required frequency of monitoring defined, the counters and logs will hopefully offer improved accuracy and precision to evaluate evidence of intervention effectiveness in real world practice.

While aggression exposure may never be completely eliminated in health care settings, better measurement to provide more comprehensive information about aggressive events and worker exposure will aid in reducing aggression rates. Evidence on health care worker exposure to aggression suggests serious physical and psychological health consequences (Gates et al., 2006; Gerberich et al., 2004; Kansagra et al., 2008; Lanza, 1983; Nachreiner et al., 2007; Needham et al., 2005; Richter & Berger, 2006; Stone et al., 2010; Whittington et al., 1996). Required aggression measures tend to capture the most extreme events. Developing additional ways to monitor and measure rates of aggression is important to learning about successful resolution of aggressive events. The results of the pilot study suggest it is feasible for workers to use handheld aggression exposure counters to collect information on events across the aggression continuum, including those that were successfully resolved. Counts of event exposure, even in a small trial, validated the difficult work that clinical staff does that is often overlooked. While aggression exposure may never be completely eliminated in health care settings, better measurement to provide more comprehensive information about aggressive events and worker exposure will aid in reducing aggression rates. Such a reduction can possibly reduce both negative patient outcomes, such as seclusion or restraint, and negative health and work outcomes for health care workers.

Acknowledgements: This publication was supported by CTSA Grant Number KL2 RR024138 or TL1 RR024137 (as appropriate) from the National Center for Research Resources (NCRR) and the National Center for Advancing Translational Science (NCATS), components of the National Institutes of Health (NIH), and NIH roadmap for Medical Research. Its contents are solely the responsibility of the authors and do not necessarily represent the official view of NIH.

The authors thank Lawrence Scahill and Susan Busch for their valuable contributions to the study. We appreciate the contributions of Rachel Mauro and Lauren Fiola to data collection and data entry and the assistance and support of the clinicians who participated in the study.

Authors

Joanne DeSanto Iennaco PhD, PMHNP-BC, APRN
E-mail: Joanne.iennaco@yale.edu

Joanne DeSanto Iennaco PhD, PMHNP-BC, APRN is an Associate Professor of Nursing at Yale University School of Nursing. Her research focuses on occupational mental health. Her most recent project, The Aggression Exposure Study, is designed to better understand the effects of aggression exposure on clinical workers.

Jane Dixon, PhD
E-mail: jane.dixon@yale.edu

Jane Dixon, PhD, is a Professor at Yale University School of Nursing, where she teaches research methods and measurement to graduate students in nursing. She is an experienced investigator with research interests in the intersection of behavior, environment and health – especially in relation to health promotion, and self and family management of illness conditions. She often focuses on problems of measurement.

Robin Whittemore, PhD, APRN, FAAN
E-mail: robin.whittemore@yale.edu

Robin Whittemore, PhD, APRN, FAAN, is an Associate Professor at Yale School of Nursing. She has expertise in behavioral interventions for youth and adults with type 1 and type 2 diabetes. Relevant to this pilot study, she has expertise in mixed methods research.

Len Bowers, RMN, PhD
E-mail: len.bowers@kcl.ac.uk

Len Bowers is a Professor of Psychiatric Nursing at the Institute of Psychiatry in London, UK. He has been conducting research in the field of violence and aggression for many years, holding many grants for large scale studies and authoring many papers.

© 2013 OJIN: The Online Journal of Issues in Nursing
Article published January 31, 2013