The sudden outbreak of the COVID-19 pandemic was an unexpected public health crisis and posed dramatic challenges to healthcare system around the world (1). Nurses were called to fight against the COVID-19 pandemic at the frontline. In this silent yet brutal battle, nurses experienced fears of infections, unexpected changes in the type of work, exhaustion due to heavy workloads, feelings of inadequate support, and uncertainty about disease control (2–4). Furthermore, they also experienced unprecedented psychological impact from the deaths of patients they cared for (5). Both the physical and mental challenges during the COVID-19 pandemic expose bedside nurses to an extremely high risk of suffering from compassion fatigue.

“Compassion fatigue” was first introduced in 1992 to describe nurses’ burnout and it has received attention from the healthcare industry since then (6). According to Figley (7), compassion fatigue is defined as the secondary trauma experienced by helpers when providing assistance to recipients with empathy as a premise, which leads to a decreased interest and ability in empathizing with oneself, as well as a sense of fatigue from the work of helping. This sense of fatigue can not only alter the original work values of helpers, but also is accompanied by a range of psychological discomfort symptoms. Compassion fatigue can result in poor health outcomes of nurses, such as insomnia, reduced immunity, headaches, sleep disorders, etc. It can also impact nurses’ performance at work, such as reduced quality of care, missed care, and medical errors (8–10). During the COVID-19 pandemic, research has focused primarily on investigating the current state of compassion fatigue (i.e., prevalence, associated factors) and the need to improve compassion fatigue in healthcare workers (i.e., neutralizing emotional trauma, utilizing relaxation strategies, improving capacity to respond to disasters and teaching coping strategies for dealing with psychological problems) (11, 12). However, few studies have explored the mechanisms by which compassion fatigue occurs during the COVID-19 pandemic. In addition, as an important healthcare workforce, nurses bear a heavy burden in clinical care and public prevention (13). Therefore, it is necessary to explore the mechanism of compassion fatigue among nurses to promote their physical and mental health.

To date, compassion fatigue has been linked to work-related stress as a main risk factor for developing compassion fatigue. One common source of work-related stress is the roles that arise from perceived multiple expectations within a social structure (14). When the expectations are conflicting, ambiguous, or overloaded, individuals experience role stress (14). Due to long working hours, heavy workload, fear of infecting their relatives and concerns about seeing patients die, nurses working during the pandemic experienced high role stress (3, 4, 15, 16). Indeed, studies have demonstrated that role stress has a strong connection with compassion fatigue (8, 17). In a study of Spanish medical staff, Ruiz-Fernández (18) and his team found that higher role stress was likely accompanied with more compassion fatigue. However, the strength of such relationship is unclear among nurses during the COVID-19 pandemic.

Recently, a series of studies have been conducted to identify potentially protective factors of compassion fatigue. Among many explored factors, psychosocial resources have drawn attention to the research community regarding their roles in compassion fatigue development. Psychosocial resources represent individual and relational factors with intrinsic values to promote physical and mental health (19). Positive psychosocial resources can help individuals recover quickly after experiencing negative events, especially major trauma, difficulties, setbacks, difficulties, stress, and even life-threatening situations (13). Psychosocial resources can effectively alleviate the effect of physical and psychological distress on compassion fatigue (13, 20, 21).

Perceived social support is one type of psychosocial resources. It is an essential component of a person’s social support system and can be defined as an emotional experience of individuals who believe they are respected, supported, and understood in society and receive support from others (22). Perceived social support is widely referred as a protective factor for psychosocial adjustment. It is believed that the establishment of social support platforms may effectively buffer the impact of the COVID-19 pandemic on the mental health of healthcare workers (3, 22). For instance, Polish nurses who had higher levels of perceived social support tended to have lower levels of compassion fatigue (23). Social support also helps individuals better manage consequent emotional damage in stressful situations (23–25).

Another psychosocial resource is emotional regulation efficacy, which can be defined as an individual’s confidence in his or her competence to regulate emotions. It can reduce emotional tension, relieve anxiety and depression, and promote mental health (26). The ability to effectively regulate emotions may alleviate the potential adverse effects when people encounter difficulties in public health emergencies (27). Appropriate emotional regulation abilities were found to reduce compassion fatigue in Iranian nurses in the COVID-19 outbreak (28). Adaptive emotion regulation strategies can buffer negative feelings caused by stress and enhance personal well-being (29). Similarly, individuals who were more able to manage their emotions perceived lower levels of stress which is a critical risk factor of compassion fatigue (30). Thus, the two psychosocial resources, perceived social support and emotional regulation efficacy, appeared to mediate the relationship between nurse role stress and compassion fatigue. However, it is unclear to what extent such complex relationships hold among Chinese nurses in public health emergencies particularly like the COVID-19 pandemic.

This study was aimed to examine the relationships between role stress, two psychosocial resources factors (perceived social support and emotional regulation efficacy), and compassion fatigue during the COVID-19 pandemic. We hypothesized that role stress was positively and directly associated with compassion fatigue (H1); perceived social support mediated the relationship between role stress and compassion fatigue (H2); The relationship between role stress and compassion fatigue was mediated by emotional regulation efficacy (H3); and perceived social support and emotional regulation efficacy exert chain-mediating roles in the influence of role stress on compassion fatigue (H4). These hypotheses are depicted in Figure 1.

This study was carried out with adherence to the STROBE guidelines after it had been approved by the institutional review board (Approval: HBZY2020-C14-01) and in accordance with the Declaration of Helsinki (31).

Using data from a cross-sectional survey, this study examined influencing factors of compassion fatigue among nurses in China during a public health emergency (i.e., the COVID-19 pandemic) and tested the possible mediating roles of perceived social support and emotional regulation efficacy between role stress and compassion fatigue. Our findings provided new insights into the complex relationship between role stress, psychosocial resources (i.e., perceived social support and emotional regulation efficacy), and compassion fatigue. These findings are informative for developing a healthy and competency nursing workforce in China beyond the COVID-19 Pandemic.

Nurses in China experienced unprecedently high compassion fatigue during the COVID-19 pandemic. In our study, more than 9 in 10 nurses reported compassion fatigue, including 65.8% with moderate or severe compassion fatigue. Previously, Hunsaker et al. (43) found that 65.9% of nurses experienced compassion fatigue in a study of 1000 emergency department nurses in the United States. Wang et al. (44) investigated 1123 clinical nurses and found that 88.38% of nurses suffered from compassion fatigue in 2020. Our findings and these results suggest that the occurrence of compassion fatigue is extremely common among nurses. Furthermore, compared to the studies before the COVID-19 pandemic, the rate of compassion fatigue is higher in our study, which suggest that nurses’ compassion fatigue has been exacerbated by the high-intensity, high-risk, high-stress work associated with the COVID-19 pandemic (1). Such a high compassion fatigue is alarming for a healthy and stable nursing workforce as other researchers have indicated that nurses are more likely to leave nursing when they experience compassion fatigue (45). Hospital executives as well as health policymakers should make a commitment to help nurses manage compassion fatigue. Research is also warranted to examine potential long-term consequence of COVID-19 pandemic on the health of frontline nurses, including compassion fatigue and other health issues.

Our study further confirms the positive relationship between role stress and compassion fatigue and provides empirical evidence quantifying this relationship in frontline nurses in China during the public health emergency. Previously, Yates (46) has concluded that job stress results in burnout, one dimension of compassion fatigue. Similarly, Alharbi et al. (17) noted that the COVID-19 outbreak increased nurses’ work stress, which raised the risk of compassion fatigue. The above observations were in line with the results we obtained. Additional to their heavy patient care duties, frontline nurses often have may other roles in life, such as caregivers, parents, husbands or wives, etc., which can add extra stress when they already face tremendous stress working at the frontline battling the pandemic as a direct care provider. Consequently, they are more likely to experience burnout, depression, and anxiety, which are closely tied to compassion fatigue.

The research findings indicate that in addition to the direct impact, role stress also indirectly influences compassion fatigue through perceived social support. According to the stress-buffering theory, high levels of social support can protect individuals from the potential negative effects of stressors (47). This suggests that an increased level of perceived social support is beneficial in alleviating the impact of stress on compassion fatigue. Wu et al. (48) conducted a cross-sectional survey involving 1464 Chinese bank employees and found that perceived social support could mediate the relationship between job stress and burnout. Additionally, through an investigation of 444 Chinese nurses, Liu and Aungsuroch (49) demonstrated that perceived social support acted as a mediator between work stress and burnout. The above results were consistent with the present study. Therefore, in highly stressful clinical environments, hospital managers could implement training activities aimed at enhancing nurses’ perception of social support to reduce their compassion fatigue.

In addition to the mediating role of perceived social support, emotional regulation efficacy also partially mediates the relationship between role stress and compassion fatigue. This finding is in agreement with previous studies. For instance, in a cross-sectional survey of 264 Italian professionals working with refugees, Tessitore et al. (50) found that emotion regulation exerted a mediating effect on the association between secondary traumatic stress and burnout. Similarly, Kshtriya et al. (51) revealed an indirect mechanistic pathway suggesting that the mediator, an emotion regulation strategy that expressed inhibition, increased the negative impact of occupational stress on the psychological well-being of emergency personnel when it was enhanced. The current research provides support for the mediating role of emotional regulation efficacy, which may be explained using a person-by-situation approach to the emotion regulation model. According to this model, individual reconstruction of emotions can lead to better mental health outcomes when exposed to high-stress environments that are relatively uncontrollable (52). This suggests that hospital administrators need to provide education and training for nurses to enhance the regulation of emotional abilities and alleviate potential adverse psychological health outcomes.

This research offered new understanding that the chain mediating effect of perceived social support and emotional regulation efficacy on the relationship between role stress and compassion fatigue. In this study, nurses were exposed to high-pressure clinical environments due to their heavy workloads, caring for critically ill patients, and performing duties in multiple roles. However, perceived social support and emotional regulation efficacy are positive psychosocial qualities acting as internal protective factors for individuals (22, 29). Researchers believed that individuals with a higher perception of social support can typically regulate their emotions better (53), which may reduce the effects of stress on compassion fatigue. When individuals face stress in their personal or professional lives, positive psychosocial resources may enhance their psychosocial resilience (53, 54). This resilience empowers individuals with courage and perseverance to confront negative thoughts generated by these stressors (13, 55). Therefore, strengthening positive psychosocial resources is crucial for maintaining psychological well-being. This pathway, which we have found, precisely explains the effectiveness of these psychological interventions during the COVID-19 pandemic. This also points to the need to deeply investigate the mechanisms and scientifically set up interventions. In the future, we need to dig deeper into positive psychosocial resources to broaden our understanding of the mechanism of compassion fatigue.

Additional to the aforementioned factors, our study also found sleep time was an important factor significantly associated with compassion fatigue. Similar to Stewart et al. (56), compassion fatigue could occur when nurses did not get enough sleep. Studies have shown that sleep deficit may cause a range of physiological stress responses that alter the composition of the gut microbiota. The altered microbiota may affect the function of the nervous and immune systems, thereby reducing an individual’s ability to cope with psychological and physical stress and making him or her more susceptible to stressful life events (57). It suggests that managers should arrange shifts scientifically and mobilize human resources reasonably, which is beneficial for nurses’ physical and mental health.

Findings from this study have several implications for nursing management, research, and policies. According to our results, perceived social support and emotional regulation efficacy have a positive protective effect against compassion fatigue. Therefore, nursing managers need to strengthen nurses’ construction of social support comprehension and enhance their emotional regulation abilities to alleviate compassion fatigue. On one hand, nursing managers could enhance nurses’ perception of harmonious interpersonal environments through some methods such as mindfulness interventions (58) and peer support programs (59). The enhancement of this comprehension of social support in interpersonal interactions helps shift nurses’ negative perceptions of stress to a positive focus on stress, thereby improving nurses’ well-being. On the other hand, nursing managers could enhance nurses’ emotional regulation abilities through activities like Yoga (60) and Balint group exercises (61), which include expressing positive feelings and managing negative emotions. Simultaneously, nurses can seek social support from family, friends, and organizations to maintain stable emotions and expand positive psychosocial resources.

This study enriched positive psychology and deeply explored the mechanism of compassion fatigue from the perspective of positive psychosocial resources. It turns out that perceived social support and emotional regulation efficacy play an essential role in buffering compassion fatigue, which provides a meaningful reference for psychological intervention for nurses in the future.

Despite the aforementioned strengths, this study has several limitations that should be noted. First, this study is a cross-sectional survey study, therefore our findings are limited to correlation relationship rather than causal relationships. It thus is necessary to conduct longitudinal studies in the future to explore the causal relationship between role stress, psychosocial resources, and compassion fatigue. Second, we only surveyed nurses in hospitals in Hubei Province. Given the geographic diversity in China, our findings may not be generalizable to nurses in other parts of China. Large-scale at national level should be considered in the future.

Nurses experience high level of compassion fatigue, a condition that is widespread and requires more attention. This study used structural equation modeling and found that when nurses are in high role stress circumstances, perceived social support and emotional regulation efficacy may buffer the negative effects of role stress on compassion fatigue. Moreover, it indicates that by building a excellent support system and managing emotions effectively, people can enhance their psychological well-being and activate positive forces within themselves. This will help nurses deal with role stress associated with the outbreak and improve compassion fatigue. It also helps to provide thoughts on promoting the mental health of nurses when uncertain public health event occurs in the future.

The datasets presented in this article are not readily available. However, the data that support the findings of this study are available from the corresponding author upon reasonable request. Requests to access the datasets should be directed to HH, hhjiabei@hbtcm.edu.cn.

The studies involving humans were approved by Hubei University of Chinese Medicine (Approval: HBZY2020-C14-01).

YZ: Formal analysis, Writing – original draft, Writing – review & editing. HH: Conceptualization, Methodology, Writing – review & editing. CY: Writing – review & editing. XW: Investigation, Writing – review & editing, Data curation. JL: Formal analysis, Validation, Writing – original draft. JX: Data curation, Investigation, Writing – review & editing. BF: Data curation, Investigation, Writing – review & editing. YC: Data curation, Formal analysis, Writing – original draft. CM: Conceptualization, Writing – review & editing.