Violence towards health care workers in a Public Health Care Facility in Italy: a repeated cross-sectional study

The problem of aggression towards health care staff is global and on the increase [1, 2]. It is difficult to gage the extent of the problem since under-reporting of violent incidents is common [3] and is probably influenced by social roles or cultural factors [4]. In fact, workers often expect violence to be part of their job [5]. However, it is also possible that people who are under considerable stress, exaggerate their responses in an effort to provide suggestions for increasing safety at work. Annual rates of physical aggression against health care workers range from a low 3.1 % [6], to a prevailing 11-25 % [7–12], or even higher (35-71 %) levels [13–16]. Non-physical aggression rates are even more difficult to evaluate; assessments range from 38 % to 90 % in a one-year period [6, 7, 9, 10, 14–18]. Studies suggest that even if patients, their relatives and friends are the main perpetrators, much of the violence encountered by health care workers is from co-workers and managers [19]. Experiences of non-physical and physical violence among health care workers are associated with decreased job satisfaction, increased occupational strain, and poor patient care outcomes [14, 20, 21]. In addition, workplace violence has a negative impact upon health care workers’ commitment to their facility [22]. Moreover, the consequences for the patients and the entire facility are serious since the perception of violence is related to adverse patient outcomes on account of the negative quality of care and treatment [10].

It is difficult to analyze the literature on health care violence because of differences in definitions and methods [23]. There are many different definitions of “violence” and even more ways of collecting data, that range from self-reports to secondary analysis of workers' compensation claims. Although many studies have examined psychiatric settings [12, 13] or emergency departments [24, 25], few have considered public health care facilities and, more specifically, variations in professions and locations [26]. Some studies focus solely on nurses and do not take into consideration other health care workers. Other studies have been carried out by external researchers with a presumably limited knowledge of the workplace. Data have been collected at a single point in time, in many cases with a low response rate (e.g.: 24 % [7], 25 % [27], 33 % [17], 36 % [8], 39 % [15]). This lack of knowledge hampers preventive measures. Furthermore, some studies refer only to patient and visitor aggressive behavior [16] and fail to consider worker-on-worker (“internal”) violence, although it has been observed that rates for worker-on-worker violence may sometimes exceed rates for patient-to-worker violence [27]. Co-worker violence threatens the well-being of hospital employees and should be regularly tracked with other forms of workplace aggression [26].

Another issue which needs to be clarified is the relationship between violence and organizational justice, a psychosocial variable which has emerged in recent years as a determinant of workers’ health [28]. A previous study indicated that verbal violence is associated with low organizational justice in nursing students [19]. Psychosocial correlates of violence, such as mental health [29], job strain, and organizational justice [28] are particularly important: firstly, as they demonstrate that workplace violence is a significant occupational hazard, and secondly, because preventive programs should primarily focus on these issues.

The Italian Ministry of Health recently issued a Recommendation specifically calling for the prevention of violence in health care facilities [30]. However, this recommendation has been disregarded, and Italian health care institutions still lack policies, strategies and administrative or behavioral provisions to counteract workplace violence.

The purpose of this study is: (1) to evaluate the prevalence (one-year incidence) of physical and non-physical violence in an Italian local health care facility from 2005 to 2009; (2) to study the association between violence and psychosocial variables in order to promote the adoption of preventive measures. To overcome the aforementioned epidemiological problems we chose to administer questionnaires during periodical medical examinations at the workplace.

Nurses and doctors were the professional groups most exposed to physical aggression and threats. A comparison with all other workers indicated that both nurses and physicians were at greater risk of physical and non-physical violence (Table 3). However, after adjustment for department type, the odds ratios were found to be lower and no longer significant (Table 4).

Results showed that workers in psychiatric and emergency services are those at greatest risk of physical aggression, since about half of all violence of this kind is concentrated in these health care areas. Psychiatric service workers reported 40 episodes of physical violence (annual rate of aggression for employed persons = 36.4 %). If the risk of a worker engaged in business services (laboratory, radiology, offices) is taken to be one, the risk of physical violence for employees in mental health services is twenty-two times higher (OR 22.7, 95%CI 9.7-52.8) and the risk of being threatened as much as twenty-seven times higher (OR 27.8 95%CI 11.5-67.5). Twenty cases were reported (annual rate = 11.4 %) in the Accident and Emergency Department. Staff in accident and emergency departments were at a significantly higher risk of violence (OR 5.1, 95%CI 2.1-12.4 for physical aggression, OR 7.6, 95%CI 3.2-18.2 for threats). The remaining episodes of violence were reported by workers in medical and surgical wards (31 cases, 8,0 %), in outpatients (9 cases, 4.3 %) and services (7 cases, 2.5 %). Workers in surgical and medical wards reported a higher risk of violence (OR 3.4, CI95% 1.5-7.9 for physical aggression, OR 2.5 CI95% 1.1-6.1 for threats) than service workers (Table 3). These associations were slightly less apparent after correction for confounders, while increases in the odds ratios continued to be highly significant (Table 4).

Non-physical aggression was more frequent than the physical type; a total of 144 workers (31.8 %) reported having been exposed to non-physical violence and described their experience. Patients (52, 36 %), their relatives and friends (35, 24 %) were responsible for more than half of these episodes; most of the remaining cases of violence were perpetrated by colleagues or superiors (50, 45 %).

The consequences of physical and non-physical violence included anger, disappointment, anxiety, distress and the intention to move to another place of work or to perform professional duties in a different way.

The occurrence of workplace violence was severely under-reported to health or police authorities. Even though we asked for a description of the worst episode of violence experienced, only about half of these incidents had been reported either to a colleague or a friend. Only in exceptional cases was a report made to the Accident Department or the police (Table 5).

A significant proportion of our sample reported experiences of workplace violence. We found that employees who experienced verbal violence in clinical settings had lower levels of perceived organizational justice and social support, higher levels of work-related stress and higher psychological problem scores than other HCWs. An important finding that emerged from this study is that isolated workers (with low social support) are exposed to violence, and that employees who experience violence are psychologically distressed. They manifest high levels of job strain and perceive their health care facility to be unfair. We also observed an interesting association between lifetime exposure to internal non-physical violence and low job control. Perceived control is a measure of power in the organization and it is significant that verbal violence, which is an abuse of power, is exerted by colleagues on workers with less authority. The cross-sectional nature of our study prevents us from indicating a definite cause for the association between violence and psychosocial variables. However, the association of workplace violence with high demand and psychological disorders, and the protective role of job control, social support and organizational justice suggest that preventive programs should target these variables.

The repercussions resulting from violence in the workplace are important as they can lead to a deterioration in staff health. A systematic review of studies on aggression showed that despite differences in countries, cultures, research designs and settings, the responses of health care workers to patient aggression are similar and include immediate responses such as frustration, fear, anger or anxiety [9, 43]. These responses may also extend to become addiction [44], burnout [45], post-traumatic stress disorder, guilt, self-blame, and shame [46], an intention to quit nursing and an intention to change institution [11]. These psychological effects can persist for months or years after the original event occurred [9]. Our study confirms that non-physical violence is associated with high work-related stress, high psychological distress and the perception of unfairness on the part of the health care facility/organization. The most likely interpretation of this result is that violence causes stress and a perception of injustice, but the cross-sectional nature of the study does not exclude the theory (suggested by some authors [14, 22]) that stressed workers are more likely to be victims of violence.

As expected, nurses were the category most exposed to physical violence. Nurses are more likely to encounter aggressive behavior because of the amount of time they spend providing direct patient care. Studies show that the chances of suffering physical violence are 7.2 and 9.0 times greater for healthcare workers with moderate and high patient contact, respectively, than for those with little or no contact [47]. In our sample, the percentage of nurses reporting one episode of physical aggression over the previous year (11.5 %) roughly corresponds to the average one-year aggression rate found by Hodgson et al. in 142 American hospitals [8], and by Gerberich et al. [9] in the Minnesota Nurses’ study on 6,300 randomly selected nurses. Our study shows that physicians are also at a significant risk of aggression. Contrary to what is normally seen, in some cases the rate of aggression was higher for doctors than for nurses. This may be due either to the role and decision-making power of doctors, especially in some high-risk medical services, or to the fact that nurses tend to report incidents less frequently than doctors.

Respondents from the psychiatric department experienced the highest overall level of patient-initiated aggression. One out of three workers reported being assaulted in the previous year, and almost three quarters of these workers had experienced physical violence during their working life. The A&E (accident and emergency) department also seemed to be a particular target for physical and verbal aggression. More than half of all the reported cases of physical and verbal violence had occurred in these high-risk departments. A high prevalence of violence in psychiatric settings has been observed in earlier studies [48] and confirmed in subsequent ones [12, 13, 27, 49, 50]. Both patient complaints originating from environmental conditions and poor communication [51] and staff-related factors, such as low work experience [12, 48], low general health [50], anxiety [13] and low job satisfaction [47, 48], are associated with aggression. The incidence of workplace violence in the emergency department has been well documented in numerous published studies [49–55]. Besides nurses and physicians, other professionals working in psychiatric [27, 56] and emergency departments [57] are also at risk.

Our study has several limitations. Firstly, since our investigation was limited to a single health care district, we cannot extend our findings to all Italian health care services. However, our results are in agreement with the literature, and we have no evidence to suggest that the situation is different in other health care facilities.

Secondly, the survey was a retrospective one, with the usual limitations of inaccurate recall of past events and of possible contamination by current events. However, the repeated measurement of aggression rates over time and their relative stability indicate that the phenomenon exists. The present study had a high participation rate compared to other similar studies on this topic, thereby increasing our confidence in the results.

Thirdly, this study was limited to an exploration of abuse from a worker perspective; no attempt was made to validate the episodes reported. In fact, no objective criteria for misconduct were specified, and the findings presented here should be considered hypothetical. What workers perceived as misconduct may not constitute misconduct in an objectively ethical or legal sense. However, it is the perception of the event and not the event itself that may have the greatest impact on the individual [58]. It has also been suggested that it may be in the interest of some professionals and labor unions to highlight rates of violence in the workplace as a way of enhancing their role in protecting workers [10]. By informing workers of the opportunity to fill in a VIF during their periodical medical examination, we chose a method that was unlikely to be influenced by the interests of consultants or unions.

Fourthly, the present study relied on self-reported measurements, which could lead to problems associated with an inflation of the strength of relationships and with the common method variance. However, to minimize problems with self-reports we used well-known validated questionnaires that have shown good reliability. Moreover, although we attempted to exclude all possible confounders, we cannot rule out the existence of residual confounding.

Our study has some strong points. The collection of data on the part of a physician who was not a direct employee of the health care facility, but who had a direct knowledge of workplaces and had had a long-lasting relationship with workers, increased the response rate and reduced the likelihood of inappropriate responses. Furthermore, this method could encourage workers to take part in prevention by suggesting possible remedies for workplace violence.

We fully agree with the suggestion of the Italian Ministry of Health that an assessment of violence is necessary in all health care organizations. No paramount episode of violence had been reported to the occupational physician before workers began completing the questionnaires.

In conclusion, our study corroborates previous reports of frequent physical and verbal aggression towards HCWs. In a standard public health service, nurses and doctors working in psychiatric services and accident and emergency departments are the most exposed to attacks. Abused workers suffer higher job stress, greater psychological distress, have a greater sense of injustice and lower social support than other workers. It is likely that a mutual increase occurs between workplace violence and psychosocial problems.

Violence at work is a hidden phenomenon and in most Italian health care facilities there is no policy for prevention. Our study indicates that the problem exists and that prevention is essential.

Most of the physical aggression and a significant proportion of the verbal aggression experienced by clinical staff are the result of patient interactions and generally regard clinical issues arising from patient care. Training workers in good working practices and alternative methods of resolving disputes is generally seen as the way to reduce the likelihood of this type of aggression [8], especially if it is accompanied by organizational and environmental safety measures. Teamwork and a supportive workplace have also been shown to mitigate this type of workplace violence [59]. The use of interdisciplinary multi-level prevention programs that have proven effective in a small private health care facility [36] would also be beneficial in public health care settings.

Our experience indicates, however, that a significant proportion of the violence encountered in the clinical setting is perpetrated by other health care workers and it is this latter form of violence that is most closely linked to psychosocial variables. Traditional methods, such as the development of personal safety skills and de-escalation techniques, or institutional policies and environmental design may not be sufficient to prevent this kind of violent behavior. These should therefore be integrated with specific intervention targeted at root causes such as conflict in the workplace. The association of violence with psychosocial variables indicates the need for far-reaching changes in health care organization that should include decision-making procedures, work climate and support, and relations between workers. Counteracting violence requires strong commitment on the part of both workers and management.

No funding was received by the first author for the study or preparation of the manuscript. Tarja Heponiemi was funded by the Academy of Finland (project no: 128002).