Intended for healthcare professionals


Workplace bullying in NHS community trust: staff questionnaire survey

BMJ 1999; 318 doi: (Published 23 January 1999) Cite this as: BMJ 1999;318:228
  1. Lyn Quine, reader in health psychology (L.Quine{at}
  1. Centre for Research in Health Behaviour, Department of Psychology, University of Kent at Canterbury, Canterbury CT2 7NP
  • Accepted 28 October 1999


Objectives: To determine the prevalence of workplace bullying in an NHS community trust; to examine the association between bullying and occupational health outcomes; and to investigate the relation between support at work and bullying.

Design: Questionnaire survey.

Setting: NHS community trust in the south east of England.

Subjects: Trust employees.

Main outcome measures: Measures included a 20item inventory of bullying behaviours designed for the study, the job induced stress scale, the hospital anxiety and depression scale, the overall job satisfaction scale, the support at work scale, and the propensity to leave scale.

Results: 1100 employees returned questionnaires—a response rate of 70%. 421(38%) employees reported experiencing one or more types of bullying in the previous year. 460(42%) had witnessed the bullying of others. When bullying occurred it was most likely to be by a manager. Two thirds of the victims of bullying had tried to take action when the bullying occurred, but most were dissatisfied with the outcome. Staff who had been bullied had significantly lower levels of job satisfaction (mean 10.5(SD 2.7) v 12.2(2.3), P<0.001) and higher levels of job induced stress (mean 22.5(SD 6.1) v 16.9(5.8), P<0.001), depression (8% (33) v 1% (7), P<0.001), anxiety (30% (125) v 9% (60), P<0.001), and intention to leave the job (8.5(2.9) v 7.0(2.7), P<0.001). Support at work seemed to protect people from some of the damaging effects of bullying.

Conclusions: Bullying is a serious problem. Setting up systems for supporting staff and for dealing with interpersonal conflict may have benefits for both employers and staff.

Key messages

  • 38% of staff in a community NHS trust reported being subjected to bullying behaviours in the workplace in the previous year and 42% had witnessed the bullying of others

  • Staff who had been bullied had lower levels of job satisfaction and higher levels of job induced stress, depression, anxiety, and intention to leave

  • Support at work may be able to protect people from some of the damaging effects of bullying

  • Employers should have policies and procedures that comprehensively address the issue of workplace bullying


Bullying in the workplace has been recognised as an important issue by trade unions in Britain for about five years. Several reports have graphically illustrated the pain, mental distress, physical illness, and career damage suffered by victims of bullying,14 but academic study began only recently.57 The most developed research comes from Scandinavia,812 where there is strong public awareness, government funded research, and established anti-bullying legislation.

Bullying presents considerable methodological problems for researchers. A central difficulty is that of definition as no clear consensus exists on what constitutes adult bullying. Although physical bullying is rarely reported, the workplace presents opportunities for a wide range of intimidating tactics. Rayner and Hoel provide five categories of bullying behaviour.7 These are threat to professional status (for example, belittling opinion, public professional humiliation, accusation of lack of effort); threat to personal standing (for example, name calling, insults, teasing); isolation (for example, preventing access to opportunities such as training, withholding information); overwork (for example, undue pressure to produce work, impossible deadlines, unnecessary disruptions); and destabilisation (for example, failure to give credit when due, meaningless tasks, removal of responsibility, shifting of goal posts).

Most definitions of workplace bullying share three elements that are influenced by case law definitions in the related areas of racial and sexual harassment. Firstly, bullying is defined in terms of its effect on the recipient not the intention of the bully. Thus it is subject to variations in personal perceptions. Secondly, there must be a negative effect on the victim. 7 8 Lyons and colleagues use the following definition: “persistent, offensive, abusive, intimidating, malicious or insulting behaviour, abuse of power or unfair penal sanctions, which makes the recipient feel upset, threatened, humiliated or vulnerable, which undermines their self-confidence and which may cause them to suffer stress.” 13 Thirdly, the bullying behaviour must be persistent.12

There have been three main approaches to research into workplace bullying. The first has been qualitative and individualistic in perspective, identifying a role for the individual in terms of vulnerability to bullying or a propensity to bully 5 14 15 and elucidating the dynamics of bully-victim relationships. The second approach is descriptive and epidemiological and is usually based on self report. 6 9 10 These studies document the prevalence of workplace bullying, the types experienced, age and sex differences, who is told, what action is taken, etc. The third approach is influenced by theories and constructs in organisational psychology and has focused on the interaction between the individual and the organisation and how aspects of the organisational structure and climate of the workplace may encourage the development of a bullying culture. 11 12

This study is a survey of workplace bullying in an NHS community trust. The objectives were to determine the prevalence of workplace bullying in the trust; to examine the association between bullying and occupational health outcomes; and to investigate the relation between support at work and bullying.

Subjects and methods

In 1996,as part of a larger survey of working life, a community NHS trust in south east England commissioned an examination of the prevalence of workplace bullying. The trust provides a range of mental health, learning disability, primary care, and child health services comprising residential care, multidisciplinary community and day service teams, health promotion, health visiting, school and community nursing services, occupational therapy, physiotherapy, speech and language therapy, and child and family psychiatry services.

We sent a questionnaire to all 1580trust employees, together with a covering letter explaining the purpose of the research and a prepaid envelope. The questionnaire was anonymous to encourage participation. A reminder was sent after three weeks, a second questionnaire after a further three weeks, and then a final reminder. The data were then entered on to computer and analysed with SPSS for Unix.


The first section of the questionnaire collected information about the participant's job, qualifications, professional group, hours worked, and supervisory responsibilities. The second contained several widely used scales to measure occupational health outcomes: job induced stress,16 job satisfaction,17 propensity to leave,18 anxiety and depression,19 and a scale to measure support at work.20 The job induced stress scale contains seven items intended to measure the existence of tensions and pressures growing out of job requirements. The job satisfaction scale uses five items to tap a worker's general affective reaction to the job.17 The propensity to leave scale provides a three item index of employees' intention to leave their job.18 The hospital anxiety and depression scale has 14items, seven of which measure anxiety and seven depression.19 Cut off points are provided to give the best separation between non-cases (0-7), doubtful cases (8-10), and cases (>) of clinical anxiety and depression.

Support at work was measured by a scale adapted from Payne.20 Workers were asked to rate on a five point scale various resources in the work environment, including feedback and support from peers and managers, access to community resources, level of workplace morale, positive working practices, and physical work environment.

The third section of the questionnaire contained questions about staff experience of bullying in the trust and examined its consequences. To avoid some of the methodological problems inherent in previous definitions of bullying we separated out the experience of bullying behaviours from their effects. Twenty types of bullying behaviour were taken from the literature, 21 22 representing each of the categories defined by Rayner and Hoel.7 Staff were asked to indicate by a yes/no response whether they had been persistently subjected to any of these behaviours in the past 12months. The final section of the questionnaire asked for sociodemographic information—age, sex, educational level—and contained questions concerning smoking and drinking habits.

Table 1

Characteristics of study participants

View this table:

Statistical analysis

A supportive work environment has been suggested to act as a coping strategy or moderator, buffering the individual from the damaging effects of work stressors such as bullying.20 The moderator effect is typically shown as an interaction term in analysis of variance.23 To test whether support at work could act as a moderator, five two-way analyses of variance were conducted. The dependent variables were job satisfaction, propensity to leave, job induced stress, and anxiety and depression. The independent variables were scores on the support at work scale, which was split at the mid point to give two groups, staff with poor support and staff with good support and scores on the bullying variable, which was divided into reported bullying and no reported bullying.


We received 1100completed questionnaires, which represented a response rate of 70%. Table 1 shows the profile of the participants. Checks made with the personnel department indicated that the sample accurately reflected the profile of the trust in terms of age, sex, and occupational groups.

Table 2 shows the proportion of staff reporting each type and category of bullying. Overall, 421(38%) people reported experiencing one or more types of bullying and 241(22%) described an incident in the past three months; 460(42%) had witnessed the bullying of others.

Table 2

Respondents reporting each type and category of bullying

View this table:
Table 3

Characteristics of the victims of bullying

View this table:

Table 3 shows the differences in occupational group, sex, age, and hours between those who reported bullying and those who did not, and table 4 shows the percentage of staff in each occupational group reporting each category of bullying. Of those reporting an incident in the past three months, 161(67%) had tried to take action about the bullying when it occurred, but most (119(74%)) were not satisfied with the outcome. Only 14/241 (6%) people had used the staff stress counselling service, which was comparatively new.


Effect of support at work on occupational health outcomes

Table 4

Number (percentage) of staff reporting each category of bullying

View this table:

The most common bully was a senior manager or line manager (128/239 (54%)), although in 80(34%) cases it was someone of the same level of seniority as the victim and in 31(12%) it was someone less senior. In 137(57%) cases the bully was the same sex as the victim, and in 54(8%) cases it was someone of the opposite sex. In 59(27%) cases the bully was male and in 144(65%) female; both sexes were involved in the remaining cases. Of the 205cases for which information was given on age, in 100(49%) the bully was older than the victim, in 57(28%) both parties were of similar age, and in 48(23%) the bully was younger.

Relations between bullying and occupational health outcomes were examined by t test or χ2 test where appropriate. Staff who had experienced bullying in the past year reported significantly lower levels of job satisfaction than other workers (mean 10.5(SD 2.7) v 12.2(2.3), t(1, 1098)=10.7, P<0.001). Additionally they had significantly higher levels of job induced stress (22.5(6.1) v 16.9(5.8), t(1, 1098)=14.4, P<0.001) and higher scores on the propensity to leave scale (8.5(2.9) v 7.0(2.7), t(1, 1098)=8.72, P<0.001) than those who had not been bullied. They were significantly more likely to suffer clinical levels of anxiety (30% (125) v 9% (60), df=1, χ2=79.3, P<0.001) and depression (8% (33) v 1% (7), df=1, χ2=32.5, P<0.001). Sixty one people who had experienced bullying reported that their health had been affected and 20had taken time off work. Altogether 335days were reported lost. Forty four per cent (56/122) of smokers who had experienced bullying reported an increase in their smoking in the previous year, and 20% (70/373) of drinkers reported an increase in their alcohol consumption.

For the analyses of variance to test whether support at work could moderate the effects of bullying 209(19%) staff were classified as having poor support and 884(81%) good support from the support at work variable. The figure shows that there were main effects of support for each outcome variable except job induced stress, and interaction (moderating) effects for three of the five outcome variables.


In this study more than a third of staff (38%) reported experiencing one or more types of bullying in the past year, which should be a cause for concern. This prevalence compares with 1in 8reporting being bullied in the past five years in a recent study of 1000workers by the Institute of Personnel and Development,24 about 1in 5found in a recent Unison survey,4 1in 3in a Royal College of Nursing Survey,25 and 1in 2of 1137mature students at one English university reporting being bullied “at some time during their working life.”6 Comparisons should be treated with caution because of differences in definition and time. The most frequently reported bullying behaviours were shifting the goal posts, withholding necessary information, undue pressure to produce work, and freezing out, ignoring, or excluding. If these four bullying behaviours only were included, the prevalence would be 32% (346). Unqualified residential care staff were most likely to report experiencing each of the five categories of bullying except enforced overwork.

Staff who had experienced bullying reported lower levels of job satisfaction and higher levels of job induced stress. They were more likely to be clinically anxious and depressed and were more likely to report wanting to leave. Three explanations could account for these associations. Firstly, being bullied leads to psychological ill health and reduced job satisfaction. Secondly, certain staff may be more likely to report being bullied than others. These may be people who are more pessimistic in outlook. Such people might also report higher levels of job dissatisfaction, propensity to leave, etc than other workers. Thirdly, being depressed, stressed, or anxious may cause a person to be bullied by unscrupulous workers who choose weaker people as their victims. Anxiety and depression may also weaken a person's ability to cope with stressors such as bullying or make them more likely to perceive other people's behaviour as hostile and critical. Longitudinal data are required to try to disentangle these effects.

Despite these difficulties, the finding that 42% of staff had witnessed the bullying of others, including many who did not report being bullied themselves, confirms that it is not purely a subjective phenomenon. Similarly, the finding that higher proportions of unqualified residential care staff and younger staff report being bullied suggests a role for aspects of the organisational climate.

Support at work

The results are consistent with the hypothesis that a supportive work environment can protect people from some of the harmful effects of bullying. Support at work may function as a buffer against stress by providing resources to enable them to cope with stressful.26 Other factors may also be able to perform this protective role, among them high levels of job control and personal dispositions such as hardiness, optimism, or sense of personal control (self efficacy). These merit further research interest.

The survey achieved a 70% response rate, which is highly satisfactory for studies of this kind.27 Checks indicated that the sample matched the profile of the trust adequately, minimising potential bias resulting from non-response. The Health and Safety Executive acknowledge in their guide Stress at Work28 that bullying is a workplace risk and advise that employers should have effective systems for dealing with interpersonal conflict, bullying, and racial and sexual harassment, including agreed grievance procedures and proper investigation of complaints. The results of this study indicate that providing a positive work environment with appropriate attention to staff support structures may be an additional way to protect people's health and welfare.


Contributors: LQ is the sole contributor.

Funding: The study was supported by a grant from the NHS trust that commissioned the research.

Conflict of interest: None.