Introduction

The COVID-19 outbreak at the end of 2019 was a huge challenge and threat for the entire healthcare industry. Nurses, in particular, were exposed to unprecedented stress and risk in the fight against COVID-19 as they spent the most time in direct contact with patients infected with COVID-19 (Lai et al., 2020). According to the International Council of Nurses, more than 2,200 nurses have been confirmed to have died during COVID-19, and an increasing number of them are leaving their jobs, with an estimated shortage of 14 million nurses by 2030 (ICN, 2021; Raso et al., 2021). The prevalence of COVID-19 means that more nurses are needed to provide specialist healthcare. Therefore, it is crucial to ensure the occupational safety and health of nurses, which may reduce the loss of nurses to a certain extent.

As one of the major occupational hazards in the healthcare industry (Awan et al., 2021), suicide is divided into suicide ideation, attempts, and death. Suicide ideation refers to the individual having the idea of dying, but not taking any actual action. Suicide attempt refers to the individual not only having the idea of death but also engaging in actual suicidal behaviors; however, these suicide behaviors do not cause death. Suicide death means that the individual has engaged in suicidal behaviors that led to death (Chinese National Health Commission, 2008; CDC, 2015). Suicide ideation and attempts are strong predictors of suicide death (Jobes & Joiner, 2019). Numerous studies have shown that nurses are at a higher risk of suicide ideation and attempts than the general population (Liu et al., 2018; Sofer, 2018). A study in Guangdong, China, found that 43.8% of nurses in general hospitals had considered or attempted suicide, and 28.7% had considered suicide at least once in the past year (Wang, 2009). A previous study showed that approximately 6.5% (n = 306) of 4,692 Chinese nurses surveyed had experienced suicide ideations during the COVID-19 outbreak (Hong et al., 2021). Another study reviewed local media and medical websites and found 46 cases of nurse suicide deaths reported or published between 2007 and 2016 (Zeng et al., 2018). This means that there is an urgent need to identify risk factors for suicide ideation and attempts in nurses and to develop targeted interventions to prevent their suicidal deaths.

Workplace bullying is another occupational hazard that threatens the safety and health of nurses. Nurses topped the list of professions reported by the International Labour Organization to be most vulnerable to workplace bullying (ILO, 2017). A recent meta-analysis of 13 articles including 5,745 nurses showed that the pooled prevalence of workplace bullying among nurses was 33.08% (Zhang et al., 2022). Individuals experiencing workplace bullying are often repeatedly isolated, intimidated, belittled, gossiped about, unfairly treated, and/or deprived of certain organizational resources by their colleagues, superiors, or subordinates (Leymann, 1996; Cleary et al., 2010; Li & Gao, 2022). Furthermore, bullying has generally been deemed as a type of psychological rather than physical violence and is categorized into four types: isolation from work, attack on professional status, attack on personality, and direct negative behaviors (Dilek & Aytolan, 2008). It is crucial to note that workplace bullying is associated with a range of adverse outcomes for nurses, such as leaving a job and experiencing psychological problems (Bambi et al., 2019; Al Muharraq et al., 2022). However, few studies have examined whether and how workplace bullying could increase suicide ideation and attempts among nurses. Thus, it is of utmost importance to determine the relationship between workplace bullying and suicide ideation and attempts among nurses.

In Australia, a recent study conducted with middle-aged workers showed that workplace bullying was associated with an increased risk of suicidal thoughts (Leach et al., 2020). In Norway, a survey-based study showed that workplace bullying was associated with subsequent suicidal ideation (Nielsen et al., 2015). These studies provide preliminary support for exploring the relationship between workplace bullying and suicidal ideation in nurses. However, they have left a few key research questions unanswered. First, previous studies (of which there are few) only adjusted for basic sociodemographic factors and work-related characteristics, with no adjustments for physical and mental health conditions known to have an impact on suicide; these include sleep problems, somatic subhealth, anxiety, and depressive symptoms (Sareen et al., 2005; Bernert et al., 2015; Racine, 2018; Ribeiro et al., 2018). This lack of focus on covariates that are well-known to influence the variables of interest may have led to biased findings. Second, from literature reviews, and consistent with the results of a systematic review (Leach et al., 2017), research on the relationship between workplace bullying and suicide attempts is lacking. Despite this lack of related studies, research has shown that suicide attempts rather than suicide ideation are more predictive of suicide deaths (Klonsky et al., 2017), warranting research on examining the strength of the relationships of suicide ideation and suicide attempts with suicide deaths. Third, few studies examined the differences in the associations between various subtypes of workplace bullying and suicide ideation and attempts. This prevents psychologists from acknowledging whether and how specific workplace bullying subtypes are greater risk factors for suicide ideation and attempts than others. Therefore, research is needed to examine the independent effects of workplace bullying subtypes on suicide ideation and attempts among nurses. Fourth, research shows that when there is bullying in the workplace, it is common for different subtypes of bullying to occur concomitantly; for example, a study showed that nurses may experience the cumulative effects of various types of bullying (Wilson, 2016). To date, no study has examined whether nurses who experience multiple types of bullying are more likely to engage in suicide ideation and attempts than those who experience a single type of bullying.

This study aimed to examine the associations between workplace bullying and suicide ideation and attempts in a large and relatively representative sample of nurses from Shandong Province, China, while adjusting for the maximum possible number of potential confounding factors. Furthermore, we aimed to discuss the independent and cumulative effects of different subtypes of workplace bullying on suicide ideation and attempts. We hope this study will help hospital managers identify nurses most at risk of experiencing suicide ideation and attempts, as well as provide new ideas for psychologists to develop targeted prevention and interventions.

Method

Participants

This study was part of the Health Longitudinal Survey of Nurses in Shandong Province, China, which was established in 2018 using a multi-stage stratified cluster sampling method. First, we stratified the cities in Shandong Province into four categories according to their urbanization development levels (Bian et al., 2009). Thereafter, we selected one city from each category using random number tables and determined four cities for the sampling procedures. Second, we used the same method to choose one tertiary hospital from each selected city as the study site; in total, four tertiary hospitals were selected. Third, after selecting the department as a sampling unit, two-thirds of the departments in each selected hospital were randomly selected from internal, external, obstetrics, and pediatrics. Fourth, all nurses in the selected departments were invited to participate in the study.

The inclusion criteria were nurses who held vocational qualification certificates and actively practiced during the investigation. The exclusion criteria were nurses who were on leave, went to other hospitals for further training, had a history of mental health problems, were currently receiving psychotherapy, or were taking antipsychotic medication.

The data used in this study were extracted during the third wave of the COVID-19 pandemic in China. Owing to the pandemic, data were collected online, using “Wenjuanxing” (a platform for electronic questionnaires, https://www.wjx.cn) and WeChat (a social media app). Data collection occurred from October to December 2020. All participants provided informed consent prior to their participation.

Measures

Study Covariates

The sociodemographic and occupational variables were selected according to previous studies (Wang et al., 2020; Zhang et al., 2020), and they included age, sex, marital status, body mass index (BMI), pregnancy, education, monthly income, years of nursing practice, type of nursing role, and number of night shifts per month.

We also measured and controlled for factors known to potentially influence the development of suicide ideation and attempt, including somatic subhealth symptoms, sleep problems, anxiety symptoms, and depression symptoms. All of these variables were coded as dichotomous variables.

The Somatic Complaints of Subhealth Status Questionnaire (Han et al., 2007) was used for measuring somatic subhealth symptoms. It consists of 16 items and is rated on a five-point scale from 0 to 4, with higher scores representing greater severity. A total score ≥ 45 was regarded as the subhealth boundary score. In this study, Cronbach’s α of the total scale was 0.95.

The Pittsburgh Sleep Quality Index Scale (Liu et al., 1996) was used for measuring sleep problems. It has 18 self-assessment items, on a scale from 0 to 3. A higher total score indicates poorer sleep quality. A total score of > 7 shows the existence of sleep problems. In this study, Cronbach’s α was 0.89.

Anxiety symptoms were measured by the Generalized Anxiety Scale (He et al., 2010). It consists of seven items on a four-point scale from 0 to 3. Higher scores indicate greater anxiety symptoms. In this study, 10 points were used as a cut-off point, and Cronbach’s α was 0.93.

Depressive symptoms were measured by the eight-item Patient Health Questionnaire (without the suicide ideation item) (Kroenke et al., 2009). It consists of eight items on a four-point scale from 0 to 3. Higher scores indicate greater depression symptoms. In this study, 10 points were used as a cut-off point, and Cronbach’s α was 0.92.

Workplace Bullying

The Chinese version of the Workplace Psychologically Violent Behaviors Instrument was used to assess nurses’ perceptions of workplace bullying inflicted on them by their managers, coworkers, and/or subordinates over the past 12 months (Xu, 2018). This 32-item scale comprises four dimensions: individuals’ isolation from work (10 items), attack on professional status (nine items), attack on personality (seven items), and direct negative behaviors (six items). Each item is scored from 0 to 5 (never happened–always). The score of each dimension is the sum of the corresponding items, and the scores of the four dimensions are added to obtain the total score. Based on prior classification criteria (Dilek & Aytolan, 2008), we dichotomized the average total scores for this scale (total score/32) into no (< 1) and yes (≥ 1). Similarly, the scores for the four dimensions were dichotomized, and participants with an average score ≥ 1 for a given dimension were considered to have suffered from that bullying subtype. In this study, the Cronbach’s α of individuals’ isolation from work, attack on professional status, attack on personality, direct negative behaviors, and the total scale were 0.95, 0.97, 0.96, 0.96, and 0.98, respectively.

Suicide Ideation and Attempts

To assess suicide ideation and attempts, we used two items from a five-item scale compiled by Chinese scholars (Tang, 2015). Three items were excluded because they assess an individual’s previous history (including the past year) of suicide ideation and attempt. Since we assessed workplace bullying that occurred in the past year, we selected one item (“have you ever seriously considered ending your own life in the past year?”) to assess suicide ideation and another (“have you ever performed a behavior to end your own life in the past year?”) to assess suicide attempts. Each item was scored from 1 to 3 (never–always). We classified participants who did not engage in suicide ideation and attempts with a score of 1 (never), and those who did with 2–3 (sometimes, always) (Tang, 2015; Zhu, 2018).

Statistical Analyses

Continuous variables were described as mean ± standard deviation (M ± SD). Categorical variables were presented as frequencies (n) and percentages (%). We used SPSS, version 26.0 to conduct all statistical analyses, and a two-sided p value of 0.05 was considered statistically significant. First, we analyzed the associations of all covariables with suicide ideation and attempts using binary logistic regression. Thereafter, in subsequent logistic regressions with workplace bullying as independent variables and suicide ideation and attempts as dependent variables, we adjusted for those covariates that showed statistically significant associations with the two dependent variables.

We first explored the associations of workplace bullying with suicide ideation and attempt separately, without adjusting for any covariates. Subsequently, we used hierarchical logistic regressions to adjust for covariates and examine the relationships between workplace bullying and suicide ideation and attempts. Regression block 1 included the following independent variables: exposures to workplace bullying (Model 1), different subtypes of workplace bullying (Model 2), and the number of bullying subtypes (Model 3). In Model 2, to explore whether specific bullying subtypes were associated with suicide ideation and attempts, we included individuals’ isolation from work, attack on professional status, attack on personality, and direct negative behaviors. In Model 3, to examine the cumulative effects of different workplace bullying subtypes on suicide ideation and attempts, we categorized nurses into five groups: those who experienced (1) no bullying (reference), (2) one bullying subtype, (3) two bullying subtypes, (4) three bullying subtypes, and (5) four bullying subtypes.

In the regression analysis with suicidal ideation as the dependent variable, we adjusted for somatic subhealth symptoms, sleep problems, anxiety symptoms, and depressive symptoms. For suicide attempts, we adjusted for mean monthly income, somatic subhealth symptoms, sleep problems, anxiety symptoms, and depressive symptoms. We reported odds ratios (ORs) and 95% confidence intervals (CIs) for all logistic regression models.

Results

Sample Characteristics and Their Associations with Suicide Ideation and Attempts

In total, we analyzed the data of 1,901 nurses, among which 96.3% were female, and their mean age was 31.31 ± 6.04 years. Among the covariates, type of nursing role was significantly associated with suicide ideation, and monthly income with suicide attempts. Furthermore, somatic subhealth symptoms, sleep problems, anxiety, and depression were all significantly associated with an increased risk of suicide ideation. Significant associations with suicide attempts were also detected. More details are provided in Table 1.

Table 1 Sample characteristics and their associations with suicide ideation and attempt
Full size table

Prevalence of Workplace Bullying and Suicide Ideation and Attempts

As shown in Table 2, 582 (30.6%) participants reported experiencing workplace bullying, 320 (16.8%) reported engaging in suicidal ideation, and 206 (10.8%) reported suicide attempts in the past year.

Table 2 Prevalence of workplace bullying and suicide ideation and attempt
Full size table

Associations Between Workplace Bullying and Suicide Ideation and Attempts

Nurses who experienced workplace bullying had a significantly higher risk of suicide ideation and suicide attempts. After adjusting for covariates, nurses with exposures to workplace bullying had a significantly higher risk of suicide ideation and suicide attempts than those who had not experienced workplace bullying (Tables 3 and 4).

Table 3 Associations between workplace bullying and suicide ideation
Full size table
Table 4 Associations between workplace bullying and suicide attempt
Full size table

Associations Between Workplace Bullying Subtypes and Suicide Ideation and Attempts

As seen in Table 3, we observed that isolation from work and attack on personality were significant risk factors of suicidal ideation. However, after adjusting for covariates, individuals’ isolation from work and direct negative behaviors were significant risk factors of suicide ideation.

As shown in Table 4, isolation from work, attack on personality, and direct negative behaviors had significant relationships with suicide attempts. After adjusting for covariates, these three variables remained significant.

Associations Between Cumulative Workplace Bullying and Suicide Ideation and Attempts

As shown in Table 3, nurses who suffered from one, two, three, and four bullying subtypes all had increased odds of engaging in suicidal ideation compared with those who did not suffer from workplace bullying. The more subtypes of workplace bullying experienced by nurses, the greater the risk of suicidal ideation. After adjusting for covariates, we observed similar results, although the risk was slightly reduced, and suffering from one bullying subtype was not a significant risk factor of suicidal ideation.

Nurses who suffered from one, two, three, and four bullying subtypes all had a significantly increased risk of engaging in suicide attempts compared with those who did not experience workplace bullying. After adjusting for covariates, only experiencing one bullying subtype became a non-significant risk factor of suicide attempts (Table 4).

Discussion

To the best of our knowledge, this was the first study to analyze the independent and cumulative effects of different subtypes of workplace bullying on suicide ideation and attempts in a large and representative sample of nurses in China. Our results revealed that nurses who were victims of workplace bullying were more likely to report suicide ideation and attempts. In addition, after adjusting for various covariates, individuals’ isolation from work and direct negative behavior were risk factors for both suicide ideation and attempts, whereas attack on personality was a risk factor only for suicide attempts. Furthermore, we found that nurses who experienced multiple bullying subtypes showed an increased risk of suicide ideation and attempts. We observed a dose–response effect of the number of bullying subtypes on suicide ideation and attempts; the higher the number of bullying subtypes experienced by nurses, the higher their risk of engaging in suicide ideation and attempts.

In this study, the prevalence of workplace bullying was 30.6%; this rate is consistent with the results of a study with nurses from tertiary hospitals in Beijing, China (Zong et al., 2020). Our finding indicates that workplace bullying among nurses in China is a growing social malpractice, which requires urgent prevention and management interventions. Furthermore, we found that the prevalence of suicide ideation and attempts was 16.8% and 10.8%, respectively, in our sample. These results are much higher than the prevalence rates of 3.9% and 0.8% reported in a recent meta-analysis of suicide ideation and attempts in the general population of China (Cao et al., 2015). Our findings reaffirm that Chinese nurses are at a high risk of suicide. Furthermore, we had to consider the impact of the COVID-19 pandemic; facing a surge in workload, worry about being infected, job segregation, and incomprehension of some patients may have contributed to suicide ideation and attempts (Awan et al., 2021).

Consistent with prior research in general populations (Nielsen et al., 2015; Leach et al., 2020), our study demonstrated that nurses who experienced workplace bullying were more likely to engage in suicide ideation and attempts. Nurses exposed to persistent and uncontrolled workplace bullying may experience psychache (Nielsen et al., 2012; Chan et al., 2019). Psychache is defined as a state of emotional/psychological pain caused by humiliation, guilt, anger, loneliness, and despair owing to a blocked or unfulfilled psychological need. When this psychological pain reaches the threshold of an individual, it can lead to suicide (Shneidman, 1993). In addition, repeated exposures to workplace bullying may cause a series of neuroendocrine and immunological changes, including increased activity of the hypothalamic–pituitary–adrenal axis and sympathetic nervous system (Reader et al., 2015; Rajalingam et al., 2021). Such repeated exposures may also elevate inflammatory markers (Jacobsen et al., 2018; Rajalingam et al., 2020), which may facilitate individual engagement in suicide ideation and attempts (Ernst et al., 2009; Van Heeringen & Mann, 2014).

Regarding bullying subtypes, after adjusting for a comprehensive number of covariates, we found that individuals’ isolation from work and direct negative behaviors were significant risk factors of both suicide ideation and attempts, while attack on personality only increased the risk of suicide attempts. According to the interpersonal suicide theory (Joiner, 2005), frustrated belonging (e.g., social isolation and poor interpersonal communication) is one of the main risk factors for suicide. Furthermore, research supports the significant predictive role of interpersonal issues for suicide (Trout, 1980; Calati et al., 2019). Direct negative behaviors refer primarily to an individual being threatened, intimidated, or forced to resign or change workplaces, or having one’s personal belongings destroyed (Dilek & Aytolan, 2008). This subtype of bullying may evoke a prolonged period of fear and high tension, which can thereby accelerate the process for individuals to experience a state of entrapment; in such cases, individuals may feel defeated and think that all escape routes are blocked or that there is no hope of rescue (Paul et al., 1998; Forkmann & Teismann, 2017). This may lead one to perceive suicide as a suitable solution for ending the state of entrapment (O'Connor & Kirtley, 2018). According to the escape theory (Baumeister, 1990), suicide can be seen as the ultimate step in escaping from the self and the world. An attack on personality may result in unfavorable self-attribution during self-reflection, aggravating self-loathing (e.g., feeling incompetent, unlikable, or guilty), and misanthropy, which may weaken one’s fear of death, strengthen their belief in suicide as a viable option, and increase the likelihood of suicide attempts (Baumeister, 1990).

After adjusting for a comprehensive set of covariates, the associations of attack on personality with suicide ideation became non-significant, whereas the effects of direct negative behaviors on suicide ideation and attempts remained significant. This may be because the effects of attack on personality on suicide ideation were mediated by one or more of the covariates we analyzed. The covariates comprised physical and psychiatric symptoms (i.e., somatic subhealth symptoms, sleep problems, anxiety, and depression symptoms). Any of these health problems experienced by nurses may have masked the effects of attack on professional status and attack on personality on their suicidality.

We also found that experiencing multiple bullying subtypes was significantly associated with an increased risk of both suicide ideation and attempts. The likelihood of suicide ideation and attempts increased with the number of subtypes experienced, especially when all four bullying behaviors were experienced. This finding is consistent with the cumulative risk hypothesis, which suggests that the accumulation of multiple risk factors increases the possibility of adverse health outcomes (Rutter, 1979). To date, our study is the first to report the cumulative effects of multiple workplace bullying subtypes on suicide ideation and attempts in nurses.

The COVID-19 pandemic is recurring and ongoing, and every healthcare worker is essential. Hospitals should create a harmonious and fair working atmosphere to reduce the occurrence of workplace bullying and establish or improve the reporting and management system of workplace bullying. Additionally, as Ralph et al.’s study shows (2021), nurses should be able to avail mental health services provided by psychologists. These professional psychologists can implement targeted psychological interventions for nurses who have developed suicide ideation or attempts to prevent them from suicide death. They can also assist hospital managers to identify individuals who have been bullied in the workplace and screen them for suicide ideation and attempts, formulating preventive measures to best ensure the occupational safety and health of nurses. These measures may contribute to nurse retention, which is beneficial to the operation of the entire healthcare system.

Limitations

The study has several limitations. First, we used a cross-sectional design; therefore, we could not infer causal relationships for the variables of interest. It is recommended that future research follows evidence-based designs. Nevertheless, we adjusted for a comprehensive number of potential confounders when examining these associations to ensure that our results were as accurate as possible. Second, we used self-reported questionnaires, which may lead to recall bias and be less objective. However, suicide ideation and attempts are considered private and stigmatizing states for individuals (Klonsky et al., 2016), denoting that confidential self-reported measures may be more reliable when assessing these constructs compared with less-confidential measurement tools. Furthermore, the scale we used to measure workplace bullying was reliable and valid (Zong et al., 2020). Third, we measured both suicide ideation and attempts using only two items. Additionally, these two items have been widely applied in past research conducted across China (Wang, 2009; Tang, 2015). Despite this limitation, a standard screening scale for assessing suicide ideation and attempts is lacking (Klonsky et al., 2016). Fourth, we recruited nurses only from tertiary hospitals in China, potentially limiting the generalizability of our findings to other settings. However, it is crucial to focus on nurses in tertiary hospitals because of the higher occupational stress and skill requirements, as well as the more complex interpersonal structures presented in these hospitals compared to other levels of hospitals (Castronovo et al., 2016). Fifth, our study lacked measures of variables related to the COVID-19 pandemic, such as experience fighting the outbreak in a high-risk area, working closely with patients with COVID-19, infection of one or more family members or colleagues with COVID-19, the experience of quarantine in a hotel or at home, personal perceptions of or attitudes toward COVID-19, reduced social activities, overwork due to staffing shortages, and lack of medical necessities. Studies have shown that these variables may be associated with suicide ideation and attempts (Farooq et al., 2021; Hong et al., 2021; Liang et al., 2022). Future studies should consider these covariates when examining the association between workplace factors, such as workplace bullying, and nurse suicide to estimate their relationship more accurately.