The effect of workplace violence on depression among medical staff in China: the mediating role of interpersonal distrust

Workplace violence has been recognized globally as a serious occupational hazard in health service occupations, and existing studies have identified that workplace violence can significantly lead to depression. Interpersonal distrust, an important topic, has also been proved associated with workplace violence and depression. However, the mediating effect of interpersonal distrust has not been tested before. Results of such testing can help us to understand further the effect mechanism of workplace violence on depression. In the current study, we collected 3426 valid questionnaires based on a cross-sectional design distributed among medical staff in Chinese hospitals. Depression, workplace violence, interpersonal distrust, social support, physical diseases, and some other social-demographic variables were evaluated. SPSS macros program (PROCESS v3.3) was used to test the mediating effect of interpersonal distrust on the association between workplace violence and depression. The data analyzed in the current study demonstrated that 52.2% of medical staff had experienced workplace violence before. Experiencing verbal violence (β = 2.99, p < 0.001), experiencing physical violence (β = 3.70, p < 0.01), experiencing both kinds of violence (β = 4.84, p < 0.001), high levels of interpersonal distrust (β = 0.22, p < 0.001), working as a nurse (β = 1.10, p < 0.05), working as a manager (β =  − 1.72, p < 0.001), suffering physical disease (β = 3.35, p < 0.001), and receiving social support (β =  − 0.23, p < 0.001) were significantly associated with depression. Workplace violence had not only positive direct effects on depression, but also an indirect effect on depression through interpersonal distrust as a mediator. Interpersonal distrust can mediate the association between workplace violence and depression. Increasing interpersonal trust or reducing workplace violence would be beneficial to promoting mental health status among medical staff.


Background
Workplace violence-defined as incidents in which staff are abused, threatened or assaulted in circumstances related to their work (ILO et al. 2002)-has become one of the most critical health issues worldwide (Hanson et al. 2015;Li et al. 2017). It can occur in many industries and professions, but people in certain professions are believed to be at higher risk of encountering workplace violence (Johansen et al. 2017;Zhao et al. 2017). Hospitals have been identified as workplaces where violence occurs most often, drawing the professional and public attention (Hsieh et al. 2016). Previous studies have also identified that medical staff in hospitals are subject to a high risk of workplace violence owing to close contact with patients and their relatives under stressful circumstances, and workplace violence has been recognized globally as a serious occupational hazard in health services occupations (Hanson et al. 2015;Johansen et al. 2017;Wu et al. 2014). A study in hospitals participating in the Occupational Health Safety Network in the United States reported that the overall rate of workplace violence injury increased by 23% annually during 2012-2015 (Groenewold et al. 2018).
In China, the rate and the number of workplace violence events against doctors have also been increasing significantly over the past decade (Shi 2015;Zhao et al. 2017). Some studies demonstrated that exposure to violence in health care settings may result in tension in doctor-patient relationships, decreased job satisfaction, reduced work performance, poor quality of patient care and poor patient care outcomes Jiao et al. 2015;Magnavita and Heponiemi 2012;Sun 2017;Zhang et al. 2017). Moreover, violence in the workplace can also lead to a variety of adverse physical and mental health issues, such as occupational strain, depression, sleep disruption, injury, and poor quality of life among medical staff (Johansen et al. 2017;Wu et al. 2014).
As stated above, existing studies conducted in different countries have identified that workplace violence can significantly lead to depression (Hanson et al. 2015;Hsieh et al. 2016;Wu et al. 2014). Workplace violence, especially in the form of a direct threat to life, can result in the development of post-traumatic disorder, stress, anxiety, and depression (Zafar et al. 2016). It has been suggested that low self-esteem, poor morale, and depression can result from exposure to verbal or physical violence at work in hospitals (Hsieh et al. 2016). Several studies have documented the effects of workplace violence on health outcomes, such as depersonalization, depression, sleeplessness, poor mental health, and post-traumatic stress disorder (da Silva et al. 2015;Hanson et al. 2015). Previous studies have described increasing pessimism, anxiety, and depression as psychological reactions occurring after exposure to violence (Inoue et al. 2006). In addition, it has been demonstrated that the health effects of workplace violence can last for years after the incidents because of fear or perceived threat (Hanson et al. 2015).
In recent years, the concept of interpersonal trust has received most attention in economic and psychological literature, and it has been predominantly associated with situations of uncertainty and risk, and with expectations about future behavior and interactions (Brennan et al. 2013). Some studies have found a relationship between interpersonal trust and mental disorders such as depression (Kim et al. 2017). Moreover, it has been demonstrated that lower levels of interpersonal trust are significantly associated with greater depression (Bova et al. 2012). Prospective studies have also shown that a low level of interpersonal trust is an independent risk factor for new-onset or long-term depression (Kim et al. 2012;Kouvonen et al. 2008). Although few studies have been conducted among health workers, we have sufficient reason to assume an association between interpersonal trust and depression.
When we reviewed the association between workplace violence and interpersonal trust, previous studies demonstrated that workplace violence can lead to a deteriorating trust-based doctor-patient relationship (Kornhaber et al. 2016). Interpersonal trust, an aspect of interpersonal relationships, should also associate with workplace violence. Moreover, interpersonal conflict is an important factor associated with hospital workers, which has been identified in previous studies (Arnetz et al. 2018). As the association between and interpersonal conflict (Ruz and Tudela 2011), we also believe there is an effect of workplace violence on interpersonal trust. As discussed above, an interrelationship is possible between workplace violence and depression, with interpersonal trust as a mediator.
Among the literature on workplace violence, interpersonal trust, and depression among medical staff, most studies have focused on their prevalence and related factors or effects on workers Hanson et al. 2015;Li et al. 2017;Shi 2015), while some studies have established their correlation with one another (Fujiwara and Kawachi 2008;Kim et al. 2012;Sun 2017;Wu et al. 2014;Zafar et al. 2016). However, little is known about the correlation between workplace violence, interpersonal distrust, and depression in health care settings. Therefore, we conducted this study to explore the associations among workplace violence, interpersonal distrust, and depression after adjusting for covariates, and to identify the mediating role of interpersonal distrust in the relationship between workplace violence and depression, using a dataset of medical staff from Shandong province in China. If the relationship can be confirmed, this will help develop interventions and strategies to reduce the negative psychological outcomes associated with experiencing workplace violence.

Setting and participants
In the current study, we collected 3426 valid questionnaires (a response rate of 98.22%, 3426/3488) based on a crosssectional design in Shandong province, China. Shandong is a large province ranked second in the population (NBS 2019), and the number of health workers also ranks first in China (NHC 2019). To obtain an accurate description of the medical staff, a multiple stratified random cluster sampling method was used to select medical staff participants. First, we randomly selected one city based on the level of gross domestic product per capita in 2017, and three cities (Qingdao, Dezhou, Zaozhuang) were selected in the current study. Second, three counties or districts were randomly selected from each city. Third, one city-level and one county-level hospital were randomly selected from each city. To this point, we had selected 12 hospitals in the 3 cities (3 citylevel hospitals and 9 county-level hospitals). Next, in each city-level hospital, three inpatient areas from each department were randomly selected. For each county-level hospital, two inpatient areas from each department were randomly selected. All the medical staff working on the survey date were asked to participate in the study.

Data collection
Data were collected from November 2018 to January 2019. The questionnaire was sent to medical staff individually, and participants could complete it anonymously in their free time. Two trained postgraduate students were stationed in the hospital to answer questions and collect the questionnaires on the survey date.

Depression
The Chinese version of the Center for Epidemiologic Studies-Depression Scale (CES-D) was used to assess the level of depression symptoms in the current study (Radloff 1977). The CES-D is an extremely popular scale with good reliability and validity for evaluating depression (Fountoulakis et al. 2001;Lewinsohn et al. 1997). The Chinese version of the CES-D has also been tested in previous studies (Cheung et al. 2006). Twenty items were employed in this scale, and the responses described participants' feelings in relation to numbers of days in the past week. The possible responses were 0 (< 1 day), 1 (1-2 days), 2 (3-4 days), and 3 (5-7 days). Cronbach's alpha was 0.852 in the current study.

Workplace violence
Workplace violence was evaluated by the question "Have you ever experienced the following behavior conducted by your patients and their relations?" Participants could choose answers from verbal violence (1), physical violence (2), both (3), and none (0).

Interpersonal distrust
We used the Interpersonal Trust Scale (ITS) to estimate the level of interpersonal distrust (Wright and Tedeschi 1975). The Chinese version of the ITS has been confirmed to have sound reliability and validity (Zhao and Xiuya 2015), and it has also been employed in several psychological studies (Ziqiang 2012). It contains 25 items to measure interpersonal trust, and answers can be chosen from 1 (completely agree) to 5 (completely disagree). A higher score denotes a lower level of interpersonal trust. In the current study, Cronbach's alpha was 0.859.

Social support
In addition, we used the Multidimensional Scale of Perceived Social Support (MSPSS) to measure the level of social support in the current study (Zimet et al. 1988(Zimet et al. , 1990). Previous studies have tested the Chinese 12-item version of this scale with good validity and reliability among Chinese adolescents (Chou 2000). Answers in the scale can be chosen from 1 (strongly disagree) to 7 (strongly agree). The total score (ranging from 12 to 84) was analyzed in the present study. Internal consistency was high in the current study (α = 0.958).

Physical disease
Physical disease was evaluated by a question asking whether the medical staff had been diagnosed with any physical disease. Possible answers were yes (1) and no (0).

Social-demographic variables
Gender was measured by male (1) and female (0). Age was assessed by the date of birth of participants, and we calculated their age at the date of survey. Married status was estimated by single, married, divorced, widowed, and others. Since few subjects were divorced or widowed, we then categorized them into single, married and others. Education was evaluated by the academic degree participants had been awarded. Choices were doctor, master, bachelor, and others.

Work-related variables
We interviewed three kinds of medical staff in the current study: doctor, nurse, and medical technician. Professional title was measured by senior, vice-senior, intermediate, junior, and others. A manager was indicated by yes (1) or no (0), with the former comprising director of the hospital, vicedirector of the hospital, director of the department, vicedirector of the department, head nurse, and deputy nurse.

Statistical analysis
IBM SPSS Statistics 24.0 (Web Edition) was used to analyze the data in the current study. T tests, one-way ANOVA, or bivariate analysis was performed to analyze the factors 1 3 associated with depression among medical staff. Linear regression was conducted to examine the factors associated with depression, and dummy variables were set for the variables with multi-classification. SPSS macros program (PRO-CESS v3.3) developed by Andrew F. Hayes was used to test the mediating effect of interpersonal distrust (Hayes and Rockwood 2019). All the tests were two-tailed and a p value of ≤ 0.05 was considered statistically significant.

Results
In the current study, we interviewed a total of 3426 medical staff in Shandong province, China. A description of the sample is provided in the second column of Table 1. We found that 52.2% of medical staff had experienced workplace violence. We also analyzed the factors associated with depression (see Table 1). The results supported that depression was associated with gender (t = 10.98, p < 0.001), professional title (F = 2.74, p < 0.05), working as a manager (t = − 4.76, p < 0.001), having a physical disease (t = 7.96, p < 0.001), receiving social support (r = − 0.37, p < 0.001), interpersonal distrust (r = 0.30, p < 0.001), and workplace violence (F = 65.68, p < 0.001).
Linear regression was used to analyze the factors associated with depression, as presented in Table 2. First, we analyzed the association without interpersonal distrust in Model A. The results supported that workplace violence was associated with depression. In Model B, when adding interpersonal distrust into the regression, both workplace violence and interpersonal distrust (β = 0.22, p < 0.001) were also associated with depression. However, all the partial regression coefficients of workplace violence had decreased, which implied a mediating effect of interpersonal distrust on the association between workplace violence and depression. Other associated factors were nursing (β = 1.10, p < 0.05), managing (β = − 1.72, p < 0.001), physical disease (β = 3.35, p < 0.001) and social support (β = − 0.23, p < 0.001).
Regarding the implications of linear regression, we analyzed the mediating effect of interpersonal distrust on the association between workplace violence and depression. The results are shown in Table 3. We found that interpersonal distrust can mediate between workplace violence and depression. The findings supported a partial effect. The indirect effects can explain about 20% (1.36*0.32/2.27) of the total effect.

The main findings
This study presents critical information about the current profile of workplace violence, interpersonal distrust, depression, and their relationships among medical staff in Chinese hospitals. More than half of the medical staff were found to be experiencing workplace violence and suffering from depression, especially those who were male, nurses, or non-managers. Workplace violence had a significant direct effect on interpersonal distrust and depression, while interpersonal distrust had a significant direct effect on depression. Moreover, there was a significant indirect effect of workplace violence on depression through interpersonal distrust as a mediator. In addition, social support was a protective factor of depression, whereas physical diseases were significantly associated with a higher risk of depression. These findings are worthy of attention because workplace violence and depression can not only lead to health issues of medical staff but also negatively affect their work performance and the efficiency of health system in the long term.

The prevalence of workplace violence and depression
The prevalence of workplace violence experienced by hospital medical staff in this study was 52.2%, which was lower than that demonstrated by some previous studies in China (Duan et al. 2019;Sun 2017). The prevalence of workplace violence in hospitals varies in different studies, and it is difficult to draw comparisons between studies (Jiao et al. 2015). This may be due to cultural differences in the perception of workplace violence in different countries, or because of the   diversity of definitions, assessment scales, study designs, and settings used in different studies (Duan et al. 2019;Magnavita and Heponiemi 2012). Numerous studies have found high levels of workplace violence among hospital health workers, and it has become more frequent and severe in recent years (Li et al. 2017). This study also demonstrated that hospital medical staff were suffering from depressive symptoms. The results indicated that the risks of Chinese medical staff experiencing workplace violence and depression are very high, which may be due to the increasing conflict and decreasing trust developing between physicians and patients in China. It has been shown that poor doctor-patient relationships may increase the risk of depressive symptoms among Chinese doctors (Wang et al. 2010). The majority of victims of workplace violence subsequently became vigilant in the clinical environment and worried about personal safety while seeing patients (Yang 2019). Therefore, targeting strategies are essential to tackle the workplace violence against and depression of medical staff from the health system perspective.

The impact of workplace violence on depression
The study also found that medical staff who have experienced workplace violence were more likely to suffer from depression than were others, which is consistent with the results of previous studies (Hanson et al. 2015;Hsieh et al. 2016;Zafar et al. 2016;Zhao et al. 2017). In addition, medical staff who have encountered both physical violence and verbal violence at the workplace may suffer from the highest level of depression, followed by those experiencing physical violence and those exposed to verbal violence. When encountering workplace violence, health workers may perceive that they do not receive respect and recognition from their patients, so they begin to doubt their professional value and status (Duan et al. 2019). Eventually, this can lead to a marked decline in enthusiasm and empathy demonstrated at work, and a significant increase in anxiety and depression (Zhao et al. 2017). Therefore, building a harmonious medical environment and reducing the prevalence of workplace violence is one effective measure to alleviate or avoid the depressive symptoms of medical staff. In addition, the findings demonstrated that certain personal and professional features were significantly associated with depression among medical staff. Thus, targeted strategies should be implemented for special groups, such as males, nurses, and non-managers. Further, social support plays a vital role in alleviating depression for health workers.

The mediating effect of interpersonal distrust on workplace violence and depression
Our findings also demonstrated that interpersonal distrust played a mediating role in the impact of workplace violence on the medical staff's depression. The results revealed that the indirect effect was positive and can be considered a supplementary mediation, indicating that a partial effect of the mediating role of interpersonal distrust was established. Obviously, these findings corresponded with the results of single analysis and linear regression analysis. Similarly, previous studies have shown that social support played a role in mitigating or mediating the impact of workplace violence on job burnout and turnover intention of physicians (Courcy et al. 2019;Couto and Lawoko 2011;Duan et al. 2019). It is noteworthy that workplace violence was negatively correlated with interpersonal distrust in the current study. Perceived interpersonal trust can play a key role in alleviating the potential stress and anxiety induced by workplace violence, which will, in turn, mediate depressive symptoms. It has been indicated that peer support plays an important role in preventing the development of depression among employees experiencing workplace violence (Hsieh et al. 2016). When medical staff suffer violence in the workplace, the importance of interpersonal distrust is highlighted, and positive measures (such as encouragement and support from their colleagues and managers) should be taken in a timely manner to help staff reduce the consequence caused by workplace violence (Duan et al. 2019;Zhao et al. 2017). Moreover, hospitals should strengthen the training and management of health professionals to reduce the harm to them resulting from workplace violence (Duan et al. 2019;Hsieh et al. 2016).

Limitations
Although this study produced some significant findings, several limitations should be noted. First, the measurement of workplace violence was obtained using a self-reporting question (whether the participants had ever experienced violence at the workplace); hence, recall bias might have affected the results. Second, it was a cross-sectional study revealing the status of the research variables at a certain time; thus, the interpretation of causal relationship between the variables was limited. It would be beneficial to confirm causality using longitudinal data in future studies. Third, the sample was selected from three cities of one province, which might compromise the representativeness of the sample, so the extrapolation of conclusions at the national level could be challenged. However, we achieved a large sample size of 3426 health workers across city-level and county-level hospitals in various cities and counties. Moreover, our findings still provide new insights for managers to maintain the stability of medical staff.

Conclusions
In conclusion, there is a high prevalence of workplace violence against medical staff in city-level and county-level hospitals. Health workers encountering workplace violence are more likely to suffer from depression than are others. In our study, workplace violence not only directly affected interpersonal distrust and depression but also indirectly affected depression through interpersonal distrust as a mediator. More attention to workplace violence and depression in hospitals is warranted. The findings suggest practical implications and policy-making references for hospital management. Preventive and managing measures, such as establishing organizational mechanisms and reporting systems, and providing training programs for health professionals, are urgently needed to deal with workplace violence and depression in hospitals. Such potential strategies would be beneficial in promoting the stability of medical staff.