A growing body of studies highlighted that one vital factor affecting the turnover intention of healthcare staff is the work environment (1). Workplace violence by patients or family members has received wide attention, but negative workplace behavior within the organization has been generally ignored. Negative workplace behavior involves psychological and physical hostility within the work environment that is pervasive and far-reaching for both the organization and its members (10). It can be divided into horizontal aggression between peers and vertical aggression between a leader and follower according to the directionality of the behavior (11). Due to the potential conflicts of interest and high-pressure work environments, negative behaviors within organizations can lead to endless stress and sustained pain for medical staff, which is no less harmful in the long run than bullying or violence from patients and family members (12). The unfavorable consequences of negative workplace behavior on turnover intention are well documented in past research (13). Previous studies have found that negative workplace behavior resulted in destroyed the harmonious atmosphere of the hospital (12), hindered healthcare workers' teamwork (14), and ultimately affected the safety of patients (15).

Although there is growing evidence that workplace incivility is an important predictor of various types of turnover intention, little is known about the association between negative workplace behavior and turnover intention among healthcare workers (16). Affective events theory (AET) can provide valuable insights for understanding the relationship between negative workplace behavior and turnover intention, which assumes that the work environment indirectly alters work attitudes by influencing work events (17). Negative workplace behavior from peers or leaders can impact healthcare professionals' working attitudes increasing the probability of turnover. Therefore, we attempt to explore the following research hypothesis:

Hypothesis 1: Negative workplace behavior has a positive association with healthcare workers' turnover intention.

Job insecurity refers to an individual's beliefs about the degree to which they fear losing their current job and the necessity of maintaining employment stability (18). The degree to which workers feel their employment is in jeopardy is reflected in their level of job insecurity (19). According to Hellgren et al. (20), there are two types of work insecurity: qualitative job insecurity, which refers to feelings of loss of important employment qualities, and quantitative job insecurity, which is based on subjective evaluations of the possibility of losing one's job. Previous study found that negative workplace behavior is a risk factor for job insecurity (21). Aisha Sarwar's study also revealed the link between job insecurity linking workplace bullying and nurse's deviant work behavior (13). Moreover, some scholars confirmed that workplace violence was positively related to healthcare workers' job insecurity. Few research, nevertheless, have looked at the possible mediating mechanisms that underlie the links between negative workplace behavior and turnover intention through job insecurity.

AET assumes that the work environment indirectly alters work attitudes by influencing work events, which subsequently trigger affective experiences (17). It offers a framework for investigating how job insecurity plays a significant role in mediating the relationship between negative workplace behavior and turnover intention (17). Events that occur to people in work environments and elicit feelings are referred to as work events (such as negative workplace behavior). Affective experiences are the feelings—whether favorable or unfavorable—that arise from work-related situations (such as job insecurity). Workplace attitudes (such as turnover intention) are directly impacted by the affective experience. Wu et al. (22) tested AET in a sample of 1,517 hospital nurses in China and found work environment was related to job satisfaction and intention to leave both directly and indirectly through two mediators: workplace violence and burnout. Therefore, we hypothesize:

Hypothesis 2: Job insecurity mediates the association between negative workplace behavior and turnover intention.

Psychological resilience is a psychological resource and stable personality trait that helps people deal with and overcome failure or adversity (23). A complicated and dynamic process, resilience involves nurses developing coping mechanisms to reduce psychological discomfort and problem-solving techniques to deal with occupational adversity (24). AET holds that personal traits can regulate the relationship between work events and affective experience (17). Healthcare professionals that possess psychological resilience are more likely to handle work-related stressors and experience personal development owing to they are more likely to mobilize internal modifications and utilize external resources (25). Furthermore, previous studies confirmed that the psychological health of emergency nurses may suffer from high rates of workplace violence, whereas resilience can help them lessen negative effects (26). Therefore, we hypothesize:

Hypothesis 3: Psychological resilience moderates the association between negative workplace behavior and job insecurity such that this positive relationship is stronger when the victim's psychological resilience is low.

Based on the above literature review, the purpose of this study was to examine whether negative workplace behavior would have both direct and indirect effects on turnover intention through job insecurity and whether psychological resilience, as a positive resource, plays a moderating role in the process of negative workplace behaviors affecting job insecurity (Figure 1).

We used a quota sampling method to collect data, 1,108 healthcare workers were recruited between February and April 2023 from three tertiary hospitals in Hunan Province, China. According to data from the China Statistical Yearbook 2020, the ratio of medical personnel (doctors, nurses, pharmacists, technicians) was 1:1.52:0.12:0.13 (27). Therefore, we recruited 400 doctors, 608 nurses, 48 pharmacists, and 52 technicians. Inclusion criteria were as follows: (1) medical personnel with relevant qualifications; (2) working time of more than 3 months; (3) signed informed consent and were voluntarily willing to participate in this study. Trainees, interns, and those who could not cooperate in completing this study were excluded. The head of nursing at three hospitals sends other nurses a link to our self-administered questionnaire via Wenjuanxing (a professional online questionnaire platform in China, https://www.wjx.cn/). On the front page of self-administered questionnaires, they were then clearly informed of the goal and relevance of this study, and their participation was kindly requested.

The required sample size was calculated using the formula n=Zα/2π(1-π)/δ2. The type I error α was set as 0.05, the Z_α/2 was set as 1.96, and the absolute error δ was set as 0.03. π = 50% was set for calculation in order to ensure sufficient size for cross-sectional study (28). A sample size of 961 was derived, and a minimum sample size of 1,057 was required given a 10% of invalid responses. Thus, 1,108individuals were recruited, which was sufficient to satisfy the cross-sectional survey's sample size requirements.

Statistical analyses were performed using SPSS 22.0 and PROCESS. The correlation between the research variables was calculated using the Pearson correlation coefficient. The impact of this model was to be assessed by multiple linear regression analyses. Based on 5,000 bootstrapped samples, this study calculated the 95% bootstrap confidence intervals (CI). A significance level of p < 0.05 was used for all variables.

The study complied with acknowledged ethical guidelines, as stated in the Declaration of Helsinki. The study received ethical approval from the Behavioral Medicine and Nursing Research Ethics Review Committee of Central South University (No. E E202217). Before starting data collection, electronic informed consent was obtained.

The majority of the participants (78.1%) were female, 26–30 years (31.9%). Most of the healthcare professionals had master's degrees (41.3%), followed by bachelor's degrees (26.2%), doctor's degrees (22.0%) and diplomas (10.5%). The present professional categories of the samples were nursing (54.9%), medicine (36.1%), pharmacy (4.3%), and technic (4.7%). The rest of the general data information is detailed in Table 1.

The descriptive statistical analysis results are shown in Table 2. Negative workplace behavior is significantly positively correlated with job insecurity (r = 0.317, p < 0.01). Job insecurity was positively correlated with turnover intention (r = 0.312, p < 0.01). There was a significant positive correlation between workplace negative behavior and turnover intention (r = 0.511, p < 0.01). This provides preliminary support for the hypothesis of this study.

We performed multiple linear regression analyses to examine our hypotheses, testing the associations between negative workplace behavior, job insecurity, psychological resilience, and turnover intention, after adjusting for gender, age, marital status, education, and working year. Such variables were regarded as the control variables. Job insecurity was tested as a potential mediator of the association between negative workplace behavior and turnover intention. We tested the mediator via the PROCESS procedure for SPSS 22.0 by calculating bias-corrected 95% confidence using bootstrapping with n = 5,000 resamples. Negative workplace behavior was positively associated with job insecurity (β = 0.322, p < 0.01) and turnover intention (β = 0.251, p < 0.01); H1 and H2 were supported, as shown in Table 3. It showed that resilience significantly moderated the association between negative workplace behavior and work insecurity in Table 3 and Figure 2, so H3 was confirmed.

The current study, like all studies, is not free from limitations. First, due to its cross-sectional nature, we were unable to draw causal relationships among negative workplace behaviors, job insecurity, resilience, and turnover intention. To validate and expand upon our findings, longitudinal data are required. Second, self-reports were used to gather all data, a process that incurs unavoidable reporting or recall bias. Future research with more objective data, including hospital turnover rates, might help confirm the results. Third, the participants were recruited from three tertiary hospitals and the sample size was relatively small, which may limit the generalizability of the findings. Future studies should consider broadening the sample's geographical scope.