Workplace violence in primary hospitals and associated risk factors: A cross‐sectional study

Abstract Aim To investigate the characteristics of workplace violence at primary hospitals in Southeast China and identify associated risk factors. Design A cross‐sectional survey design was used for this work. Methods We distributed a workplace violence questionnaire among medical staff at primary hospitals in Southeast Zhejiang Province, China. The data were collected between December 2016 and December 2017. We analysed the categorical data by using the chi‐square test and expressed it as frequencies. The risk factors were analysed by using multiple logistic regression analysis. Results Among the 2,560 questionnaires, 1,842 (71.9%) medical staff indicated that they had experienced workplace violence. Verbal assault was the most common type, followed by physical and sexual assault. Furthermore, gender, age, marital status, education, technical position and number of hospital beds' numbers were independent risk factors.


| INTRODUC TI ON
Workplace violence (WPV) in hospitals has become a global problem (Spector et al., 2014), but the incidence thereof in Chinese hospitals is much higher than in other parts of the world (Arnetz et al., 2015).
The World Health Organization has defined that hospital WPV as the action that takes place when "a medical practitioner is insulted, threatened and attacked in the workplace poses a challenge for his safety, well-being and health" (Bowers et al., 2007). This type of violence can be physical, sexual or psychological (Campbell et al., 2011).
With the rapid growth of China's social economy, the relationship between medical staff and patients has become increasingly tense, leading to a breakdown in trust crisis between them. However, Chinese studies have not been sufficiently focussed on the characteristics of violence, related risk factors (Liu et al., 2015) or systematic research in primary hospitals. Of all the medical settings, the primary ones are the most vulnerable in China. Therefore, to prompt primary hospitals to concentrate more on WPV prevention and control, we aimed to investigate the incidence of violence and associated risk factors among medical personnel in these hospitals in China. The following workplace violence inclusion criteria were established based on the World Health Organization definition of violence in the workplace: (a) psychological, physical and/or sexual violence suffered by hospital staff on hospital premises; (b) the violence occurred in the 12 months prior to the survey; (c) the victim must be a hospital staff member; (d) the place of violence is limited to hospitals; (e) the time limit is 12 months before the survey.

| Design and setting
This cross-sectional study was conducted in four county-level primary hospitals, namely the Jiangshan, Wuyi, Yongkang and Fuyang Hospitals in Zhejiang Province.

| Participants
Based on the World Health Organization's (WHO) definition of WPV, we used the following inclusion criteria in our study: (a) hospital staff members who were victims of (b) psychological, physical and/or sexual violence, (c) on hospital premises, (d) in the 12 months prior to the survey. The data were collected from registered medical staff at county-level primary hospitals in the Southeast Zhejiang Province who had more than 1 year of work experience.

| Measures: Hospital Workplace Violence Questionnaire
The Hospital WPV Questionnaire was designed by Chen during his doctoral dissertation to study WPV among 7,198 medical workers in 20 different medical facilities. We obtained permission from the author and used the questionnaire to investigate the characteristics of, and staff responses to, WPV in the hospitals. The questionnaire includes two sections with 38 items. Part One (nine items) reflects the different types of WPV with two psychological violence (insults and threats); four physical violence (physical attacks without injuries; physical attacks with mild, visible or serious injuries); and three for sexual violence (sexual harassment or flirtation; assault; and attempted rape or rape) items. Part Two (14 items) mainly describes the victims' most severe WPV incident in the 12 months preceding the survey. Part Three (15 items) is about participants' personal data, their WPV knowledge and the hospital's response. The retest reliability of the questionnaire was 0.803. Results showed that the validity and reliability of the questionnaire were good, and we obtained consent from the original author to use the questionnaire.

| Ethical considerations
The Ethics Committee of Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University (20171120-8) approved the study.
Before we distributed the questionnaires, we assured the medical staff that the questionnaire will be used for academic research, their personal information will remain confidential, and they could withdraw at any stage. Moreover, the participants signed informed consent forms.

| Data collection
We collected the data between December 2016 and 2017 by distributing 2,700 questionnaires of which 2,560 (94.8%) valid ones were returned ( Table 1). The questionnaires were anonymized and sealed and posited by the participants into a specific box at a nursing station within 2 weeks of receiving the questionnaire.

| Data analysis
Descriptive statistics were used to analyse the demographic data and WPV incidence. Moreover, we analysed the risk factors using multiple logistic regression analysis by applying SPSS version 21 (IBM) zero statistical software package (IBM Corporation), and p < .05 was considered statistically significant.

| WPV-related risk factors
When we compared the characteristics of the patients who had experienced WPV with those who had not, we found significant differences in gender, age, marriage, education, technical position and number of hospital beds (p < .05). However, no significant differences were observed in terms of job type, employment patterns and years of work experience (

| D ISCUSS I ON
Primary hospitals are particularly vulnerable settings for WPV, with a general WPV prevalence as high as 60%. Alarmingly, we found a 4% prevalence rate over a one-year period. Furthermore, previous studies focussed on clinical settings in medium and large hospitals, but addressed limited types of WPV. That partly accounts for the relatively low prevalence of WPV. We investigated all common WPV styles and associated risk factors at several primary hospitals, which would account for a high event incidence.
Similar to previous studies (Yang et al., 2018) and consistent with regular conflicts in Chinese daily life, verbal assault was the most prevalent type of WPV, followed by physical and sexual assault.
However, even though verbal assault is common in primary medical settings, psychological violence usually does not cause serious physical harm. Considerable domestic and foreign research reported that medical staff who experience prolonged verbal assault will experience low job satisfaction (Hanson et al., 2015), and professional fatigue (Heponiemi et al., 2014), or resort to resigning (Kim et al., 2018

TA B L E 3 (Continued)
disability. Although no fatal WPV cases were found, 23 of the medical staff had suffered from severe physical violence. Notably, these results only refer to the 12-month period preceding the survey.
Sexual violence is often a "blind spot" in research. Many women do not report such incidences due to shame, or fear of being condemned or not believed (Arnetz & Arnetz, 2001 Research has shown that women are more vulnerable to attacks (Guay et al., 2016;Lipscomb et al., 2006)-most registered nurses in China are female. However, in our study, not only gender but also age, marital status, education, position and the number of hospital beds were significantly correlated with the WPV incidence in primary hospitals.
In terms of age, younger, inexperienced medical staff are not well versed in communicating with problematic patients, which results in medical disputes. The disputes seem to evolve around errors that have occurred (Shi et al., 2015) probably due to limited work experience and sense of responsibility. Notably, senior and experienced medical staff are more likely to gain patients' trust than novice ones.
This result is consistent with other research (Svoboda, 2013).
Regular workers are a high-risk WPV group. This is related to the fact that they undertake clinical work and face patients for a prolonged time. Because of this, the doctors and nurses who have most access to patients are most vulnerable to WPV (Bruns et al., 2007;Stevenson et al., 2015). In Western countries, the contact between medical staff and patients centres around employment. Therefore, it is not often that medical staff tend to the same patients perma-

| Limitations
As this exploratory study involved convenience sampling and was conducted at selected primary hospitals in the Zhejiang Province, Abbreviations: CI, confidence interval; OR, odds ratio; WPV, Workplace violence.