Horizontal violence can cause serious mental health problems for nurses, particularly anxiety, depression, and post-traumatic stress disorder. However, the intrinsic linkage mechanism between mental symptoms of anxiety, depression, and post-traumatic stress disorder in nurses exposed to horizontal violence is unclear. This study aims to elucidate the characteristics of anxiety, depression, and post-traumatic stress disorder networks among nurses with horizontal violence exposure.
Methods
Data for this cross-sectional study were obtained from the baseline portion of a short longitudinal survey conducted at four tertiary hospitals in Shandong Province, China. A total of 510 nurses with horizontal violence exposure completed the General Information Scale, the Negative Acts Questionnaire, the Seven-item Generalized Anxiety Disorder Scale, the Nine-item Patient Health Questionnaire, and the Four-item SPAN. The network model was constructed using network analysis. The expected influence and the bridge expected influence of nodes were calculated. The stability and accuracy of the network were estimated.
Results
The results show that A4 (Trouble relaxing) and P1 (Startle) had the highest expected influence in the network. D9 (Suicidality ideation) and A5 (Restlessness) were the key bridge symptoms.
Conclusions
“Trouble relaxing”, “Startle”, “Suicidality ideation”, and “Restlessness” are all mental symptoms that need to be urgently improved the most in nurses exposed to horizontal violence. Nursing administrators and policymakers should implement mental health intervention programs for these symptoms as early as possible to maximize nurses’ mental health.
Huimin Wei and Mengqi Liu contributed equally to this.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
The International Committee of the Red Cross reported that workplace violence against nurses in healthcare settings is a major and growing public health and occupational safety problem worldwide [1, 2]. The main perpetrators of workplace violence include patients, their families, supervisors, and coworkers [3]. However, internal workplace violence from coworkers (i.e., horizontal violence) results in a higher risk of poor mental health among nurses compared with external workplace violence from patients and their families [4]. Nurse-to-nurse horizontal violence (HV) is defined as aggressive, bullying, intimidating, or divisive hostile behavior that occurs between nurses, including verbal, physical, and emotional violence [5]. HV compromises the health of nurses and the safety of patients and is now increasingly recognized as a prominent and hot-button societal issue within the healthcare system [3]. Surveys have shown that the global prevalence of HV among nurses ranges from 25.3% to 96.0% [6]. Continuous and repeated exposure to traumatic events, such as violence and bullying, puts nurses at a greater risk of developing mental health disorders [3, 7]. Nurses’ impaired mental health not only affects their performance and jeopardizes patient safety [7, 8] but also leads to an increased desire to leave their position, contributing to nursing brain drain [7]. Therefore, the mental health of HV-exposed nurses needs urgent attention.
Workplace violence often triggers a variety of mental health issues. Of these, anxiety, depression, and post-traumatic stress disorder (PTSD) are three of the most globally prevalent and debilitating mental consequences of violent trauma and have been the focus of considerable research attention [9]. In a study on public health workers, Tiesman et al. [10] found that exposure to workplace violence was significantly associated with an increased likelihood of reporting anxiety, depression, PTSD, and suicide. Havaei [11] reported that emergency nurses experiencing direct or both direct and indirect dual workplace violence exposures were 2–4 times more likely to report high levels of anxiety, depression, PTSD, and burnout than nurses with no violence exposure. It is evident that workplace violence seriously undermines the psychological well-being of healthcare professionals. However, previous studies have rarely distinguished perpetrators or have focused primarily on the threat of violence from patients and family members, hindering our observation of the damaging effects of workplace violence from coworkers with greater threat potential (i.e., HV) on nurses’ mental health [12].
Anzeige
Previous research has consistently demonstrated that anxiety, depression, and PTSD are closely correlated and often comorbid [13, 14]. It is important to note that these symptoms comprise multiple sub-symptoms, and exploring the degree of association between these sub-symptoms and identifying key symptoms could help find more precise targets for treatment interventions. The traditional latent variable model assumes that symptoms arise from an underlying disease entity and are independent of each other [15]. This view ignores the interplay between symptoms; however, the associations between symptoms are common and fundamental to mental disorders. Network analysis is an emerging approach to understanding the interactions between symptoms in psychopathology. According to the network theory of psychopathology [16], the emergence of mental disorders is caused by strong causal associations and feedback loops between symptoms. Based on this approach, the active interrelationships between variables can be represented as a network of nodes (symptoms) and edges (relationships between symptoms) [17]. Among them, the most influential “central symptoms” and “bridge symptoms” are critical for triggering and maintaining mental illness networks, which may be potential targets for intervention [17]. Thus, by applying network analysis to mental symptoms, we can obtain a deeper understanding of their intricate relationships and identify specific symptoms that could be crucial to the progression of the illness.
Previous studies have examined the network characteristics of anxiety, depression, and PTSD in various groups. Jiang et al. [18] independently introduced 20 sub-symptoms of PTSD, while anxiety and depression were each included as a whole in a study of a COVID-19-exposed population. The results showed that the strongest association was between “loss of interest” and “depression,” and that the most prominent symptom was “self-destructive/reckless behavior” in the cluster of PTSD symptoms [18]. Fico et al. [19] independently introduced eight sub-symptoms of depression, while anxiety and PTSD were each included as a whole in a study of COVID-19-infected patients. The results found that “sad mood” and “lack of pleasure” were most closely linked in the depressive symptom cluster, with the most central symptom being anxiety [19]. However, these studies have several limitations. First, they have not simultaneously explored the potential associations among the microscopic symptoms of anxiety, depression, and PTSD, which can lead to a lack of a nuanced and comprehensive understanding of the complex associations among the three. Second, they have not agreed on the strength of the associations and significance of symptoms, leading to an inability to identify the most urgent targets for improvement, which greatly reduces the effectiveness of intervention programs. Third, this evidence is unclear in the HV-exposed nurse population, and an in-depth investigation of related studies is urgently required. Therefore, this study aimed to explore the deeper associations between anxiety, depression, and PTSD symptoms in HV-exposed nurses using network analysis to identify the dominant edges and core symptoms.
Methods
Study design, setting, and participants
This cross-sectional study was conducted according to the STROBE reporting guidelines. Data for this study were obtained from the baseline portion of a short longitudinal survey conducted at four tertiary hospitals in Shandong Province, China. Convenience sampling was used to select the hospitals, and nurses with HV exposure were screened as study participants. The inclusion criteria for practice nurses were as follows: (i) holding a certificate of nursing qualification from the People’s Republic of China, (ii) working in clinical nursing for six months or more (In China, it takes six months or more of work to qualify for clinical work.), and (iii) score greater than 19 as measured by the Chinese version of the Negative Acts Questionnaire (i.e., HV exposure). The exclusion criteria were as follows: (i) absent from work, vacation or further training, (ii) suffered a major accident or received psychological treatment.
Sample size
This study follows Epskamp’s recommendations for the sample size for network analysis; a total of 190 edge weights parameters (i.e., \(\frac{\text{k}\times (\text{k}-1)}{2}\), where k represents the number of nodes in the network) and 20 threshold parameters (i.e., one for each node in the network) need to be estimated when constructing a network with 20 nodes [17]. To reliably estimate a network model, the number of observations required usually exceeds the total parameter (i.e., 190 + 20 = 210 in this study) [17]. Considering that 20% were invalid questionnaires, the final required sample size was 263.
Anzeige
Data collection
Data were collected between February 2023 and March 2023 using the Wenjuanxing platform. The platform was used to configure the same cell phone or WeChat to answer only once and submit only after all questions had been answered. Every sample hospital’s nursing department was contacted to inform them of the purpose and method of the study to obtain their consent. After the request was approved, the nursing department was formally requested to deliver an electronic questionnaire via WeChat to the nurses who met the requirements. Nurses were provided with information on the purpose and method of the study. All those participating in the study completed an informed consent form. Of the 1,143 nurses who completed the questionnaire, 532 who had experienced HV were screened, with an exposure rate of 46.5%. After excluding 22 nurses who had received psychological treatment, 510 nurses were finally included in the study.
Measures
The General Information Scale
General data were collected, including information about demographic (gender, age, education level, marital status, and number of children) and work-related (years of experience, department, title, number of night shifts per month, and authorized strength).
The Negative Acts Questionnaire
The Chinese version of the Negative Acts Questionnaire developed by Li [20] was used to measure nurse-to-nurse HV over the previous six months. It comprises 19 items scored on a five-point Likert scale (1 = “never,” 5 = “once a day”). The scores for each item are summed, with higher overall scores denoting more severe HV. Cronbach’s α in this study was 0.953. In reference to previous studies [21, 22], if any one of the negative acts occurred once or more (i.e., total scores > 19), the respondents were considered to have suffered from HV.
The Seven-item Generalized Anxiety Disorder Scale (GAD-7)
Anxiety symptoms were evaluated using the Chinese version of the GAD-7 [23]. It comprises seven items scored on a four-point Likert scale (0 = “never,” 3 = “almost every day”). The scores for each item are summed, with higher overall scores denoting more severe anxiety. The cutoff scores for positive and negative anxiety were set at 10. Cronbach’s α in this study was 0.920.
The Nine-item Patient Health Questionnaire (PHQ-9)
Depressive symptoms were evaluated using the Chinese version of the PHQ-9 [24]. It comprises nine items scored on a four-point Likert scale (0 = “none,” 3 = “almost every day”). The scores for each item are summed, with higher overall scores denoting more severe depression. The cutoff scores for positive and negative depression were set at 10. Cronbach’s α in this study was 0.910.
The Four-item SPAN
PTSD symptoms were evaluated using the Chinese version of the Four-item SPAN [25], which comprises of: startle, physiologic arousal, anger, and numbness. It is scored on a five-point Likert scale (0 = “not distressing at all,” 4 = “extremely distressing”). The scores for each item are summed, with higher overall scores denoting more severe PTSD. The cutoff scores for positive and negative clinical PTSD were set at five. Cronbach’s α in this study was 0.933.
Statistical analysis
Network construction and visualization
Using the Extended Bayesian Information Criterion graphical least absolute shrinkage and selection operator [26, 27], a Gaussian graphical model of anxiety–depression–PTSD in nurses exposed to HV was constructed and visualized using the R-package qgraph of Rstudio (version 4.2.3) software [28, 29]. After adjusting for other nodes, partial correlation analysis was performed to determine the net correlation between each pair of variables [30]. In the network, the 20 sub-symptoms of anxiety, depression, and PTSD were represented by nodes, and the link between the two nodes was indicated by edges. Blue and red edges denoted positive and negative associations, respectively, with thicker edges denoting stronger correlations. The Fruchterman-Reingold algorithm was used to locate nodes with stronger correlations near the center of the network [31].
Centrality estimation
We computed three centrality indices—strength, betweenness, and closeness—to determine the significance of each symptom [32]. However, Owczarek et al. [33] argued that expected influence (EI) outperforms conventional centrality indices when a negative correlation exists in the network. Therefore, we chose EI as the main metric to determine the importance of nodes. In addition, to identify bridge symptoms in the network, we calculated bridge expected influence (BEI) using R-package networktools [34]. A higher BEI score indicated a greater chance of the current community spreading to neighboring communities [35]. Additionally, the predictability of each node was calculated using the R-package mgm to assess the extent to which a particular node was interpreted by its neighboring nodes and the controllability of the network model [36].
Stability and accuracy estimation
Three steps were used to estimate the stability and accuracy of the networks using the R-package bootnet [17]. First, the accuracy of the edge weights was evaluated by calculating 95% confidence intervals (CIs) using a non-parametric bootstrapping method based on 1000 bootstrap samples. Second, the stability of the centrality indices was evaluated by calculating the correlation stability coefficient (CS-C) using a case-dropping bootstrapping method. The network was considered stable with a CS-C value of at least 0.25, and ideally higher than 0.5 [17]. Third, differences in network properties (i.e., whether any two edge weights, EIs, or BEIs differed significantly from one another) were tested using bootstrapped difference tests.
Ethical considerations
This study was approved by the Ethics Committee of the School of Nursing and Rehabilitation, Shandong University (approval no. 2023-R-111). All participants provided informed consent, and their information was kept strictly confidential and anonymous.
Results
Sample characteristics
The mean age of the 510 nurses exposed to HV was 33.71 ± 6.47 years (range: 21–55 years). The years of experience was 11.21 ± 7.27 years. Most participants were female (n = 469, 91.96%). The prevalence of anxiety, depression, and PTSD was 12.94% (n = 66), 18.82% (n = 96), and 17.06% (n = 87), respectively. Overall, 26.27% (n = 134) of the nurses experienced at least one mental health problem. There were significant differences in positive rates of anxiety, depression, and PTSD among nurses who differed in the number of night shifts per month. Detailed information on other demographic and work characteristics is provided in Supplementary Table 1. Abbreviations, mean scores, EI, BEI and predictability for each node are listed in Table 1.
Table 1
Descriptive statistics of the items in the anxiety–depression–PTSD network
Abbreviation
Mean (SD)
EI
BEI
Predictability
A1: Nervousness
0.757(0.755)
0.946
0.128
0.786
A2: Uncontrollable worry
0.665(0.783)
1.101
0.151
0.823
A3: Excessive worry
0.778(0.770)
1.081
0.173
0.834
A4: Trouble relaxing
0.769(0.784)
1.175
0.269
0.846
A5: Restlessness
0.549(0.718)
0.878
0.297
0.723
A6: Irritability
0.771(0.758)
0.950
0.197
0.783
A7: Feeling afraid
0.551(0.729)
1.011
0.236
0.773
D1: Anhedonia
0.875(0.773)
0.936
0.074
0.724
D2: Sad mood
0.776(0.742)
1.038
0.238
0.757
D3: Sleeping
0.851(0.849)
0.749
0.187
0.617
D4: Fatigue
1.016(0.836)
1.009
0.054
0.713
D5: Appetite
0.816(0.813)
0.781
0.108
0.606
D6: Worthless
0.659(0.729)
0.932
0.127
0.700
D7: Concentration
0.602(0.718)
1.003
0.106
0.717
D8: Motor
0.533(0.710)
1.044
0.166
0.732
D9: Suicidality ideation
0.412(0.666)
0.807
0.388
0.640
P1: Startle
0.604(0.893)
1.149
0.132
0.798
P2: Physiological arousal
0.496(0.777)
0.894
0.114
0.741
P3: Anger
0.800(0.936)
0.771
0.205
0.698
P4: Numbness
0.529(0.847)
0.954
0.162
0.742
A1–A7 are GAD-7 items on anxiety; D1–D9 are PHQ-9 items on depression; P1–P4 are SPAN items on PTSD; SD Standard deviations
Network structure
The network structure of anxiety–depression–PTSD comprising 20 nodes is shown in Fig. 1. Of the 190 possible connected edges, 93 were non-zero edges (48.95%), including 87 positive and six negative edges. The mean weight of the edges was 0.051. The most strongly connected edges appeared within their respective communities rather than across communities. Within the depression community, the most significant edge was D7 (Concentration)–D8 (Motor) (weight = 0.388). Within the PTSD community, the strongest edge was P1 (Startle)–P3 (Anger) (weight = 0.384), followed by P1 (Startle)–P2 (Physiological arousal) (weight = 0.383). Within the anxiety community, A3 (Excessive worry)–A4 (Trouble relaxing) (weight = 0.348) was the strongest edge. These edges were significantly stronger than the vast majority of other edges according to bootstrapped difference tests of edge weights (Supplementary Fig. 1). The most robust transdiagnostic edge throughout the community was A7 (Feeling afraid)–D9 (Suicidality ideation) (weight = 0.187), followed by A6 (Irritability)–P3 (Anger) (weight = 0.177), A5 (Restlessness)–D8 (Motor) (weight = 0.144), and D9 (Suicidality ideation)–P4 (Numbness) (weight = 0.105). All edge weights are listed in Supplementary Table 2. The non-parametric bootstrapping results demonstrated relatively narrow 95% CIs for the edge weights, suggesting that the network was stable and accurate (Supplementary Fig. 2).
Fig. 1
Network structure of anxiety–depression–PTSD among HV-exposed nurses
×
Node centrality and stability
A4 (Trouble relaxing) (EI = 1.175) had the highest EI, followed by P1 (Startle) (EI = 1.149), indicating that they were the most core symptoms (Fig. 2). Bootstrapped difference tests showed that these nodes had significantly higher EIs than 68.42% and 52.63% of the other nodes in the network, respectively (Supplementary Fig. 3). D3 (Sleeping) (EI = 0.749) and P3 (Anger) (EI = 0.771) had the lowest EIs, suggesting that they might be borderline symptomatic in the network. The case-dropping bootstrapping results indicated that the CS-C value for EIs was 0.594 (i.e., after discarding 70% of the original data, there was a 0.594 correlation between the centrality of the remaining 30% and that of the original samples), indicating that the estimation of the node EIs was sufficiently stable (Fig. 3).
Fig. 2
EI and BEI of the network of anxiety–depression–PTSD among HV-exposed nurses
Fig. 3
The stability of EI by the case-dropping bootstrapping
×
×
Bridge nodes centrality and stability
D9 (Suicidal ideation) (BEI = 0.390) had the highest BEI, followed by A5 (Restlessness) (BEI = 0.300), suggesting that they may be key bridge symptoms contributing to the comorbid symptoms of anxiety, depression, and PTSD (Fig. 2). Bootstrapped difference tests showed that the BEIs of these nodes were significantly higher than 73.68% and 31.58% of other nodes in the network, respectively (Supplementary Fig. 4). D4 (Fatigue) (BEI = 0.054) had the lowest BEI, indicating that it was the least contagious among communities. The case-dropping bootstrapping demonstrated that the CS-C value for BEIs is 0.439 (i.e., after discarding 70% of the original data, there was a 0.439 correlation between the bridge centrality of the remaining 30% and that of the original samples), indicating that the node BEIs had an acceptable level of stability (Supplementary Fig. 5).
Node predictability
The predictability of a node is indicated by the percentage of coloring of the circle surrounding it (Fig. 1). The predictability of the nodes ranged from 0.606 to 0.846, with an average predictability of 73.77%, indicating that nearby nodes accounted for an average of 73.77% of the variance of each node. A4 (Trouble relaxing) had the highest predictability (0.846), followed by A3 (Excessive worry) (0.834), A2 (Uncontrollable worry) (0.823), and P1 (Startle) (0.798), whereas D5 (Appetite) had the lowest (0.606).
Discussion
This study is the first to explore the complex interactions between the microscopic symptoms of anxiety, depression, and PTSD in HV-exposed nurses using psychological network analysis. The findings revealed specific activation patterns in anxiety, depression, and PTSD symptoms associated with HV exposure. In addition, the present study identified two core symptoms, A4 (Trouble relaxing) and P1 (Startle), which may be crucial for triggering and sustaining the anxiety–depression–PTSD network. Finally, we identified D9 (Suicidality ideation) and A5 (Restlessness) as bridge symptoms that increased the risk and severity of comorbidity between anxiety, depression, and PTSD.
Anzeige
Previous studies have found that, in networks composed of different communities, the strongest edges generally exist within communities rather than cross-community edges [37]. The results of the network analysis in this study reaffirmed this finding. Specifically, the strongest edges were present in the depression community (i.e., D7 (Concentration)–D8 (Motor)), PTSD community (i.e., P1 (Startle)–P3 (Anger) and P1 (Startle)–P2 (Physiological arousal)), and anxiety community (i.e., A3 (Excessive worry)–A4 (Trouble relaxing)). The strong association between D7 (Concentration) and D8 (Motor) is supported by previous studies [38, 39]. A network analysis of psychiatric distress in survivors of severe COVID-19 trauma exposure [38] found more significant edge weights for D7 (Concentration)–D8 (Motor), which suggests that violent trauma and COVID-19 trauma may result in similar psychological symptomatic responses and highlights the importance of the association of “Concentration” and “Motor” in psychological consequences of trauma exposure. Another study of indirect trauma-exposed people who watched flood-related videos [39] found that the “Motor” symptom was on the shortest path between exposure to the “flood victims seeking help” video scene and the “Concentration” symptom, indicating a high strength of association between “Concentration” and “Motor”. The two were most closely linked in this study probably since nurses experiencing HV tend to have various physical and mental symptoms that affect concentration and energy [3]. Thus, they are not supported in accomplishing normal activities; conversely, lack of activities also results in depleted concentration [40], creating a vicious circle. Additionally, P1 (Startle)–P3 (Anger) and P1 (Startle)–P2 (Physiological arousal) are the edges with the second and third strongest connections in the network, respectively. Blechert et al. [41] found that the startle response and anger were significantly positively correlated in patients with PTSD and that anger increased proportionally with the startle response. Bryant et al. [42] also highlighted a close correlation between the startle response and physiological reactivity. In our study, HV was a harmful “stressor” and “trauma” for nurses, and nurses’ exaggerated startle response often implied a lowering of the body’s threshold for perceiving the threat [42]. This threat triggers overactivation of the sympathetic-adrenomedullary system and the release of catecholamines, which, in turn, provides feedback to the brain and affects the neural structures that control emotions, triggering undesirable emotions such as anger [43]. Further, the hypothalamic–pituitary–adrenal axis is also activated in overstressed states, leading to an excessive release of glucocorticoids, triggering sweating, diarrhea, and other physiological discomforts [44]. The strong connection between A3 (Excessive worry) and A4 (Trouble relaxing) has also been confirmed in previous studies [45, 46].
Furthermore, strong cross-community edges are key edges that lead to activation and contagion between different communities, which is essential to the emergence and persistence of symbiotic psychological issues. Specifically, this study identified four strong cross-community edges: A7 (Feeling afraid)–D9 (Suicidality ideation); A6 (Irritability)–P3 (Anger); A5 (Restlessness)–D8 (Motor); and D9 (Suicidality ideation)–P4 (Numbness). These findings hint the critical pathways of A7 (Feeling afraid)–D9 (Suicidality ideation) and A5 (Restlessness)–D8 (Motor) may be the underlying mechanisms for the strong association between anxiety and depression, often leading to their co-occurrence. This conclusion is very strongly supported by a previous study of COVID-19-exposed nurses [47], which found that “Feeling afraid”– “Suicidality ideation” and “Restlessness”– “Motor” were the two most powerful edges driving mutual reinforcement between anxiety and depressive symptoms. Additionally, the connective pathway between “Irritability” and “Anger” may be the key connection that activates the development of anxiety and PTSD. Irritability refers to a lesser degree of easily annoyed and impatient emotional states, which tends to be an internal experience [48], whereas anger refers to a greater degree of an intense and uncontrollable emotional outpouring of rage, which tends to be an external reaction [48]. The strong correlation between the two may be explained by the fact that an individual’s increased sensitivity to violent trauma after experiencing a traumatic event leads to a lower threshold for expressing undesirable emotions, making the individual more irritable; the long-term accumulation of irritability leads to the expression of strong anger [49]. Further, the close association between depression and PTSD may be through the linkage mechanism of more microscopic symptoms, that is, “Suicidality ideation” and “Numbness.” Previous studies have often supported the conclusion that anxiety, depression, and PTSD are comorbid, and the above four strong cross-community edges may be key intrinsic mechanisms leading to the coexistence of the three, which needs to be explored in further studies.
Centrality results indicated that A4 (Trouble relaxing) had the highest EI. In other words, A4 (Trouble relaxing) was the most central symptom in the anxiety–depression–PTSD network of HV-exposed nurses and should be considered the driver and trigger of all psychological stress symptoms. In a qualitative systematic review [50] of HV issues in nurses, nurses from multiple studies reported that they were often oppressed, ostracized, and verbally abused at work, e.g., “bloody useless”, “stupid”, which tends to put them on a high level of mental tension that making it difficult for them to relax. The strong centrality of “Trouble relaxing” is also supported by a previous study of trauma-exposed adults [13] which found that “Trouble relaxing” was highly centrally weighted and was one of the central symptoms activating the network of anxiety, depression, and PTSD symptoms. Additionally, the EI for P1 (Startle) was second only to that for A4 (Trouble relaxing), suggesting that “Startle” may be another important trigger in the network with a high potential to activate other neighboring symptoms and further lead to serious mental problems. In a qualitative meta-synthesis [51] on HV issues among nurses, a nurse noted that she wanted to report that she was being bullied at work, but was threatened that if she reported it she would be ganged up on by other nurses, which startled and frightened her. The centrality of “Startle” is confirmed by the above results on the edges, that is, P1 (Startle)–P3 (Anger) and P1 (Startle)–P2 (Physiological arousal) have the second and third strongest connections in the entire network, respectively. The sensitization model of PTSD [52] also states that traumatic experiences make individuals more sensitive to threats, which may lead to stronger startle responses. Startle is a particularly strong form of fear conditioning that can lead to excessive psychophysiological responses to stimuli [53], and thus may be prominent in the anxiety–depression–PTSD network. This study emphasizes the urgency of prioritizing screening and intervention for “Trouble relaxing” and “Startle” symptoms in nurses experiencing HV to minimize the mental health impact.
The results of bridge centrality suggested that D9 (Suicidality ideation) and A5 (Restlessness) had the highest BEI and should be considered key bridge symptoms leading to the comorbidity of anxiety, depression, and PTSD. A nurse in this study once revealed, “When I was publicly reprimanded and humiliated after I made a mistake due to my inexperience, I was even suicidal at that moment. Later, I was often so afraid of making mistakes at work that I would walk around all the time and couldn’t sit quietly.” The above results on edges hint that the bridging role of the two may be realized through the specific pathways of D9 (Suicidality ideation)–A7 (Feeling afraid), D9 (Suicidality ideation)–P4 (Numbness), and A5 (Restlessness)–D8 (Motor). This finding provides insights for nursing administrators to find labor-saving and effective intervention targets, that is, interventions targeting “Suicidality ideation” and “Restlessness” can achieve more significant double or even triple stacking effects. Specifically, when HV-exposed nurses are severely depressed, interventions targeting “D9 (Suicidality ideation)” not only help to ameliorate depression, but also simultaneously reduce the mental crises of anxiety and PTSD through intrinsic links to A7 (Feeling afraid) and P4 (Numbness), respectively. Similarly, when HV-exposed nurses are more anxious, interventions targeting “A5 (Restlessness)” not only alleviate anxiety, but also ameliorate the negative psychology of depression through potential linkage pathways with D8 (Motor). This conclusion is supported by a study of psychiatric disorders in nurses [30], which found interconnections between “Restlessness” and “Motor” to be a powerful cross-diagnostic strength in the network of anxiety and depression symptoms. Additionally, empirical clinical studies [47] have found that targeted mania treatment for “Restlessness” has significant clinical value in improving anxiety and depression.
Anzeige
Limitations
This study has some limitations. First, it was cross-sectional and exploratory; thus, it was unable to identify causal relationships or long-term dynamic changes between symptoms. Future research should conduct longitudinal cross-lagged network analysis to determine the direction between mental symptoms and provide more accurate information for interventions. Second, as HV events may be more sensitive in the nursing environment, a format based on self-report questionnaires may mask the severity of the mental damage that HV inflicts on nurses. Therefore, future studies should attempt to conduct structured clinical interviews to accurately and in-depth assess the mental health threat posed by violent trauma to nurses. Additionally, owing to the uneven gender distribution of the nursing population, the majority of participants in this study were female (91.96%), which could limit the generalizability of these findings. Future studies should include more male nurses or validate the applicability of our results in other populations.
Conclusions
To the best of our knowledge, this study is the first to explore the network structure of anxiety, depression, and PTSD in HV-exposed nurses using psychological network analysis, elucidating deeper associations among the more microscopic symptoms of the three mental health disorders. A4 (Trouble relaxing) and P1 (Startle) were identified as symptoms requiring urgent clinical intervention. Interventions targeting bridge symptoms, that is, D9 (Suicidality ideation) and A5 (Restlessness), can help to simultaneously improve the mental consequences of anxiety, depression, and PTSD.
Relevance for clinical practice
The damage caused by HV to nurses’ mental health cannot be ignored, and intervention programs are urgently needed to deal with this psychological crisis. The present study elucidated the deeper associations between anxiety, depression, and PTSD among HV-exposed nurses from a microscopic perspective, which is expected to identify more precise and effective intervention targets. Specifically, to address the two key symptoms of the anxiety community (i.e., “Trouble relaxing” and “Restlessness”), nursing managers can organize appropriate relaxation training for nurses before shifts, such as mindfulness meditation, breathing relaxation or playing some soothing music [54]. This can improve both their mental health and professional performance. To address the psychological symptoms of “Startle” in the PTSD community, nursing administrators can organize startle-relieving workshops and invite nurses to share measures to mitigate the startle response to help nurses develop optimal coping strategies [55]. In response to the “Suicidality ideation” in the depression community, nursing administrators must be vigilant. Early screen of nurses at high risk for suicide must be conducted and timely and professional psychotherapy should be provided [56]. In addition, to minimize the occurrence of HV, potential prevention strategies should be implemented as soon as possible. Healthcare organizations can conduct online surveys of HV behaviors and organize training on HV, such as role-playing scenarios to help nurses identify HV behaviors and learn possible responses [57]. Healthcare administrators can also invite experienced nurses to share HV incidents they have previously witnessed or been the target of, using examples to inform nurses of ways to address such issues [58]. In addition, nursing administrators should actively establish a supportive unit climate that encourages nurses to collaborate and learn from each other in order to reduce the breeding ground for HV behaviors [58]. Policymakers can also use this information to design learning curricula for nurses to help them prevent, recognize, and respond to HV [59].
Acknowledgements
Thank the nursing department of four hospitals and participants for their support to this study.
Anzeige
Declarations
Ethics approval and consent to participate
This study was approved by the Ethics Committee of the School of Nursing and Rehabilitation, Shandong University (approval no. 2023-R-111). All participants provided informed consent and their information was kept strictly confidential and anonymous.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
