Although in Australia the risk of death or serious physical injury from a violent workplace incident is quite remote, each year about one Australian health worker is murdered at work1 and large numbers are either verbally abused, bullied or assaulted.2

Workplace violence has been defined in various ways, and behaviours ranging from verbal abuse and threats to sexual harassment, physical assaults and homicide may be included. The precise definition adopted will affect incidence and severity rates of workplace violence. An inclusive definition developed specifically for the health sector is:

There are two core risk factors for violence in any workplace:

• face-to-face contact with clients or customers; and

• cash or high-value goods on site which may attract perpetrators of instrumental violence.4

Marked variations in risk exist between different occupational groups, reflecting the relative presence or absence of these two core risk factors.5 These patterns generally hold across different countries so that high-risk jobs in one country tend to also be high-risk jobs in others. For example, throughout the industrialised world there is a very high rate of homicide of taxi drivers, who also experience high levels of verbal abuse and assault. Conversely, white collar workers, with little face-to-face contact with members of the public, generally have very low incidence and severity rate. Fast-food outlet workers tend to experience high levels of verbal abuse but homicide is rare (apart from those located in suburbs with a high risk of hold-ups). Other occupational groups have diverse experiences essentially determined by levels of exposure to known risk factors.

In health care settings, one of the two core risk factors — face-to-face contact with patients and visitors — is particularly common. The Box gives workplace verbal abuse, bullying and assault rates for various Australian industry sectors. These estimates were collated after separate face-to-face surveys conducted with representative samples of employees. In each case, interviewees were requested to state precisely what, if any, form of workplace violence they had experienced over the previous 12-month period. While space prohibits an exposé of these different studies here, it is clear that health care ranks fairly highly compared with many of the groups.6 Almost all the cited “bullying” events were from one staff member to another (Box). A similar pattern of variable risk across industry sectors and occupational groups is also evident in Britain.4

In 2001–02, representatives from the International Labour Office, International Council of Nurses, World Health Organization and Public Services International initiated an international collaborative program to develop policies and approaches to prevent and eliminate violence in the health sector. Outcomes included:

• a series of country-specific research studies (Brazil, Bulgaria, Lebanon, Portugal, South Africa, and Thailand and a linked Australian study; see <http://www.icn.ch/sewworkplace.htm>);

• a Synthesis report of the commissioned country reports;3 and finally

• the drafting of Framework guidelines for addressing workplace violence in the health sector.7

Across the various country-specific research studies, more than 50% of health workers reported experiencing one or other form of workplace violence in the previous 12 months; ambulance officers were at greatest risk, followed by nurses. The linked Australian health study involved face-to-face interviews with a representative sample of 400 public health employees, including medical officers, nurses, allied health, ancillary and ambulance workers.2,6 About two-thirds (67%) of all interviewees said they had been verbally abused in the previous 12-month period, 10.5% had been bullied, and 12% assaulted, rarely resulting in physical injury. As in other countries, incidence rates varied between and within health occupations, with ambulance officers most at risk, followed by nurses. Among medical officers interviewed, 62% had been verbally abused, 15% bullied and 17% assaulted over this period. The perpetrators of verbal abuse and assaults were predominantly clients and visitors, however, other staff members were responsible for almost all bullying events. Nevertheless, only a small proportion (between 8% and 10%) of these events had been formally reported — providing health authorities with limited evidence on which to base preventive planning. Similar incidence rates have been reported in general medical practice,8 including the study by Magin et al in this issue of the Journal (page 352).9

Comprehensive workplace violence prevention strategies have been developed and are available.4,10-12 However, in our experience, there is a tendency among health workers to favour preventive strategies that have only been trialled at other health sites. We would encourage health workers to consider the full range of prevention strategies, many of which have been well documented and evaluated in other industry sectors — including, in particular, those published in the scientific literature for criminology and occupational health and safety (OHS).

All workplace violence prevention strategies — regardless of the setting, health care or otherwise — should be multifaceted and organisation-wide, and involve widespread consultation with all workers (including casuals and those off-site) in their development and implementation. In health care settings, this may involve nursing agencies, ambulance officers, and workers in the community and remote locations.

In the OHS “hierarchy of control” approach to violence, designing out risk is the preferred action in all workplace environments, and should include health care settings such as hospitals, clinics, general practices, and ambulances and other vehicles. For example, to remove or minimise risk, careful attention can be paid to the design of buildings (eg, placement of windows) and their fittings (eg, counter height and width, and choosing chairs for waiting areas that cannot be easily lifted and thrown).2,10,13 The least preferred action is sole reliance on staff training, as the causes of workplace violence are multifactorial and hence simplistic solutions are unlikely to be effective in prevention.

Chief executive officers (CEOs) should demonstrate commitment to zero tolerance of workplace violence. They should encourage cultural change and show enhanced concern for workers’ safety. Strong encouragement should be given to formal reporting of workplace violence, including the removal of covert penalties and the excessive filling in of forms. Regular violence vulnerability audits should be conducted by independent OHS professionals.

The files of serial perpetrator patients should be “flagged” to forewarn other staff (the best predictor of violence is past aggression), and, ultimately, such patients should be sanctioned, which, although difficult, is an essential component of a comprehensive workplace violence prevention strategy.14 Particularly in the UK National Health Service, the sanctioning of perpetrators of violence has become increasingly common; however, the deterrent reach of penalties can be limited by the lack of capacity of certain patients to control their behaviour, including those suffering from mental illness and dementia.

Under the OHS legislative framework in each Australian state and territory, primary responsibility rests with employers and CEOs to provide a safe place and a safe process of work for their employees, including those who work off-site. OHS obligations are not diminished by the rights of patients to confidentiality or to treatment. Employers (including public health departments) and CEOs may be prosecuted for a breach of this “duty of care” with respect to their employees, as well as being liable under common law.15

The core challenges for health authorities will be to:

• implement a zero-tolerance policy with effective prevention strategies encompassing all health occupational groups, and especially those working off-site and in general medical practice;

• adopt preventive strategies which have been successful in other industry sectors; and

• develop and implement a deterrent sanctioning policy for perpetrators.

The group of articles in this issue of the Journal will further assist the health industry to recognise the potential benefits from involvement with the Australian Patient Safety Foundation (APSF) database (Benveniste et al, page 348);16 identify verbal cues of imminent overt aggression in mental health settings (Forster et al, page 357);17 understand possible causes of under-reporting of violence against emergency department staff (Kennedy, page 362),18 and to also recognise the particular vulnerabilities faced by those working in general medical practice.9