What this article is about
In brief
A 2017 web survey of 850 Hong Kong nurses found that 44. 6% reported workplace violence during the previous year; verbal abuse or bullying was the most common form, and the low response rate limits the precision of the prevalence estimate.
For nursing students
Study summary
Cheung and Yip examined workplace violence among registered nurses working in Hong Kong. Members of the Association of Hong Kong Nursing Staff who had an email address were invited to complete a Chinese-language web survey during October and November 2013. Of 16,082 possible email recipients, 850 nurses participated.
The questionnaire asked about verbal abuse or bullying, physical assault, and sexual harassment during the previous 12 months. It also recorded who was responsible, how nurses responded, and how strongly an incident affected their mental health. A validated Chinese version of the DASS-21 measured recent symptoms of depression, anxiety, and stress. The researchers used weighting and logistic regression to estimate prevalence and examine factors associated with reported violence.
Overall, 44.6% of respondents reported at least one form of workplace violence. Verbal abuse or bullying was reported by 39.2%, physical assault by 22.7%, and sexual harassment by 1.1%; participants could report more than one form. Patients and patients’ relatives were named more often than colleagues or supervisors as perpetrators.
The final regression model linked reported violence with rotating shift work, job dissatisfaction, conflict with colleagues, a history of deliberate self-harm, anxiety symptoms, and clinical position. These are statistical associations, not proof that any one factor caused violence. Among nurses who experienced violence, 49.4% rated its mental-health impact as moderate and 25.5% rated it as severe.
Only 5.3% of invited members completed the survey, so respondents may differ from nurses who did not participate. The sample also contained a larger proportion of men than the wider Hong Kong nursing workforce, prompting the researchers to apply gender weighting. Because exposure and possible correlates were measured at the same time, the study cannot establish which came first.
The authors proposed accessible incident-reporting procedures, prompt managerial responses, zero-tolerance policies, and violence-management education across general and mental-health nursing programs. The survey did not test whether any prevention program reduced violence.
Original publication
Source abstract and study details
Read the source abstract
Abstract Background Nurses are especially vulnerable to violent and other forms of aggression in the workplace. Nonetheless, few population-based studies of workplace violence have been undertaken among working-age nurse professionals in Hong Kong in the last decade. Methods The study estimates the prevalence and examines the socio-economic and psychological correlates of workplace violence (WPV) among professional nurses in Hong Kong. The study uses a cross-sectional survey design. Multivariate logistic regression examines the weighted prevalence rates of WPV and its associated factors for a population of nurses. Results A total of 850 nurses participated in the study. 44.6% had experienced WPV in the preceding year. Male nurses reported more WPV than their female counterparts. The most common forms of WPV were verbal abuse/bullying (39.2%), then physical assault (22.7%) and sexual harassment (1.1%). The most common perpetrators of WPV were patients (36.6%) and their relatives (17.5%), followed by colleagues (7.7%) and supervisors (6.3%). Clinical position, shift work, job satisfaction, recent disturbances with colleagues, deliberate self-harm (DSH) and symptoms of anxiety were significantly correlated with WPV for nurses. Conclusions WPV remains a significant concern for healthcare worldwide. Hong Kong’s local health authority should put in place a raft of zero-tolerance measures to prevent WPV in healthcare settings.
Reviewed findings
Main findings
- Of 850 respondents, 379 (44.6%) reported workplace violence during the previous 12 months.
- Verbal abuse or bullying was reported by 39.2%, physical assault by 22.7%, and sexual harassment by 1.1%; respondents could report more than one form.
- Patients (36.6%) and patients’ relatives (17.5%) were reported more often as perpetrators than colleagues (7.7%) or supervisors (6.3%).
- Shift rotation had the largest adjusted association in the final model (aOR 2.67); job dissatisfaction, deliberate self-harm history, colleague conflict, anxiety symptoms, and clinical position also remained significant.
Using the findings
Practice considerations
- Violence policies and incident systems should address verbal abuse and bullying as well as physical and sexual violence.
- The authors call for a transparent, easy-to-use reporting process and prompt managerial follow-up so nurses do not treat violence as an unavoidable part of the job.
- The paper recommends extending aggression-management and de-escalation education beyond mental-health nursing to general nursing programs.
For educators
Teaching and appraisal notes
This paper supports a focused lesson on prevalence surveys, composite outcomes, weighting, and causal restraint. Recruitment reached 16,082 email-registered members of a professional association, but 850 completed the survey. Learners should distinguish the achieved sample from the 5.3% participation rate and consider why weighting for gender cannot repair every form of non-response.
The workplace-violence outcome combined verbal abuse or bullying, physical assault, and sexual harassment, with participants able to select more than one category. This explains why the category percentages should not be added together. The survey paired an adapted ILO/ICN/WHO/PSI instrument with the Chinese DASS-21 and used multivariable logistic regression.
Shift rotation, job satisfaction, colleague conflict, deliberate self-harm history, anxiety symptoms, and clinical position remained statistically associated with reported violence. Neither timing nor mechanism can be resolved in a cross-sectional design. Anxiety and deliberate self-harm must not be presented as personal causes of victimization.
The clinical-position result requires source checking: the results give an adjusted odds ratio near 0.34 for front-line nurses relative to charge nurses, while the discussion describes front-line nurses as more susceptible. Readers should inspect the model coding and table before stating the direction.
Critical appraisal
Limitations
- Only 5.3% of the nurses reached by email completed the survey, leaving the prevalence estimate vulnerable to non-response and self-selection bias.
- Recruitment was limited to full-time, Chinese-reading members of one professional association who had registered email accounts.
- The cross-sectional design cannot determine whether anxiety, job dissatisfaction, or colleague conflict preceded or followed workplace violence.
- The article’s accounts of the clinical-position result are not fully consistent, so the regression table and category coding should be checked before stating its direction.
Classroom use
Discussion Questions
- Why should the three violence-category percentages not be added to calculate an overall prevalence?
- What types of non-response bias could arise when 5.3% of invited association members complete a workplace-violence survey?
- How would a longitudinal design help clarify the relationship between anxiety symptoms and workplace violence?
- What features would make an incident-reporting process usable and psychologically safe for nurses?
Source-based questions
Frequently asked questions
Why should the 44.6% prevalence estimate be interpreted cautiously?
The 850 respondents represented only 5.3% of association members reached by email. Gender weighting could not remove every possible difference between respondents and non-respondents.