Anxiety and burnout in infectious disease nurses: the role of perceived stress and resilience

Abstract

Background

Nurses serving in infectious disease ward represent a distinct occupational group that has attracted considerable attention following epidemic outbreaks. However, prior to this study, no research had delved into the underlying mechanism linking anxiety to burnout symptoms among infectious disease nurses. This study aimed to explore investigate the association between anxiety and burnout among nurses working in such environments and scrutinized the mediating role of perceived stress and the moderating influence of resilience on the principal relationship.

Methods

Employing a cross-sectional study using a web-based design, data were collected from 1,579 clinical nurses working in infectious disease ward across 50 hospitals in China. Participants responded to questionnaires assessing anxiety, perceived stress, resilience and burnout. Statistical analyses encompassed descriptive statistics, one-way analyses of variance, independent-samples t-tests, Pearson correlations, and bootstrapping techniques to evaluate the indirect and moderating effects.

Results

The study revealed that 62.13% of the respondents reported high levels of burnout, and 55.92% experienced moderate to high degrees of emotional exhaustion among infectious disease nurses. Significant correlations were observed between anxiety, perceived stress, resilience, and each component of burnout (P < 0.05). Notably, the mediating effect of perceived stress was accounting for 30.61% of the relationship between anxiety and burnout. Simple slope analysis demonstrated that perceived stress significantly predicted emotional exhaustion at both low (B = 0.854, t = 16.586, and P < 0.001) and high (B = 0.498, t = 9.503, and P < 0.001) levels of resilience. The perceived stress and emotional exhaustion were more serious when resilience levels were lower.

Conclusion

Anxiety was identified as a critical risk factor for burnout among nurses in infectious disease units. The relationship between anxiety and burnout was markedly affected by the levels of perceived stress and resilience. Specifically, the deleterious impact of perceived stress on burnout was amplified in nurses with diminished resilience relative to those with heightened resilience. Based on these findings, it is imperative to allocate resources for stress management programs and resilience training. Such initiatives would bolster support for nurses in infectious disease wards, ultimately enhancing their job satisfaction and well-being.

Introduction

Burnout, a syndrome precipitated by chronic occupational stress, manifests predominantly as emotional exhaustion, engendering a cascade of adverse health outcomes including respiratory complications, cardiovascular disorders, gastrointestinal issues, cephalalgias, type 2 diabetes, and chronic fatigue [1]. Moreover, this condition exerts profound impacts on mental health, materializing as depressive episodes, anxiety disorders, sleep disruptions, and suicidal contemplations [2, 3]. The Maslach Burnout Inventory delineates three core components: persistent emotional exhaustion (EE), depersonalization (DP), and diminished personal accomplishment (PA) [4, 5]. A meta-analysis of data from 1,986 nurses elucidated prevalence rates of 31% for elevated emotional exhaustion, 18% for increased depersonalization, and 46% for decreased personal accomplishment [6]. Within the healthcare sector, particularly amongst nursing professionals, burnout is a pervasive psychosocial phenomenon. Prolonged exposure to burnout leads to a decline in work engagement, exacerbating feelings of weariness and self-doubt [7]. Nurses grappling with burnout are predisposed to committing errors, displaying reduced compassion, and delivering substandard care, thereby jeopardizing patient safety [8, 9]. Thus, addressing burnout among nursing staff is paramount, essential for sustaining superior patient care, retaining skilled personnel, safeguarding the welfare of healthcare providers, and fostering a nurturing work environment.

Conversely, anxiety is typified by feelings of trepidation, disquietude, and strain, encompassing emotional, cognitive, and behavioral facets. While a moderate degree of stress can elicit motivation and excitement, chronic anxiety can detrimentally affect mental and physical health, alongside undermining occupational efficacy. Its impact on professional performance manifests as impaired cognitive function, diminished productivity, and an escalation in errors [10]. Healthcare practitioners enduring protracted anxiety are susceptible to burnout, with potential cascading effects into chronic stress and disrupted sleep patterns [11]. Hence, the effective management of persistent anxiety is imperative for preserving professional competence and overall wellness in healthcare professionals. Simultaneously, exacerbated anxiety is correlated with maladaptive coping strategies, such as increased substance abuse, exacerbated stress, depression, and suicidal inclinations [4]. Nurses, in particular, are at an elevated risk of experiencing high-stress levels [5]. In a recent study involving 1,807 registered nurses, investigators discerned that 43.4% manifested symptoms indicative of anxiety [12]. The authors of a study examined anxiety and related factors among 1807 registered nurses, finding that 43.4% exhibited signs of anxiety. Another research conducted corroborated these findings, showing that 37.3% of nurses suffered from anxiety symptoms, which were linked to a myriad of personal and occupational determinants [13]. These determinants consisted of marital status and household income as personal factors, and specialty and duration of employment as job-related elements. Moreover, investigations have revealed that frontline nurses have encountered escalated anxiety levels during the COVID-19 pandemic. Implementing interventions aimed at alleviating anxiety could prove instrumental in shielding nurses from adverse consequences.

The COVID-19 pandemic has had a profound impact on global economies, healthcare infrastructures, and public health frameworks [14]. It is imperative to acknowledge that the COVID-19 crisis led to significant biological and psychosocial hazards among nurses [15]. Research has identified a substantial correlation between experiences of burnout and intentions to resign within one year among 3,030 nurses during this pandemic. Over 64.3% of nurses reported experiencing burnout, while 36.5% expressed intentions to depart from their hospital employment, respectively [8]. Moreover, prior investigations have demonstrated that the heightened workload during the COVID-19 pandemic contributed to substantial burnout among nursing staff, precipitating issues such as insomnia, workplace aggression, and substance misuse [9, 16, 17]. Post-pandemic, governments worldwide are recalibrating their infectious disease control strategies to better reflect contemporary epidemiological conditions. Nurses specialized in infectious diseases bore the brunt of frontline care during the outbreak, resulting in notable burnout [18]. However, there remains a paucity of data regarding their psychological health and burnout levels in the post-COVID-19 era. Although there is mounting evidence pertaining to COVID-19-induced burnout, anxiety, and perceived stress among nurses, existing research lacks a comprehensive exploration of the interplay between burnout and anxiety concerning these variables. Consequently, this study endeavors to examine the degrees of anxiety and perceived stress among nurses specializing in infectious diseases, and to elucidate how anxiety influences stress and burnout after the COVID-19 breakout.

Perceived stress as a mediator

Anxiety can lead to burnout both directly and indirectly by influencing factors like perceived stress. Stress arises when different internal and external elements trigger the body, and this can be due to numerous physiological and mental factors [19]. Nurses are broadly acknowledged as one of the professions facing the most significant stress levels [20]. According to a study of 2,895 Iranian nurses, 78.4% reported high occupational stress levels [21]. In China, nurses face significant overwork because of the vast number of patients and limited healthcare resources. The study by Li et al. (2016) revealed 32.8% of nurses experienced anxiety symptoms, which were exacerbated by stressful events [22]. An examination of frontline nurses in Wuhan during the COVID-19 crisis found 33.4% of participants had anxiety, which was associated with perceived stress and sleep disturbances [23].

Increased levels of perceived stress were strongly linked to more serious burnout. Recent findings indicate a potential link between perceived stress and burnout. Research by Luan X and colleagues has revealed a strong positive association between stress and burnout among head nurses and senior nurses, and burnout is predicted by perceived stress [24]. A study involving 366 nurses revealed that 85.5% of female nurses faced psychological distress, and there was a positive correlation between burnout and psychological distress [25].

Resilience as a moderator

An individual’s resilience is characterized by their capacity to bounce back from situations that endanger stability, survival, and development, which is crucial for safeguarding mental well-being [21, 26]. In essence, resilience involves effectively adjusting and fostering positive changes when confronted with challenges, stress, trauma, and significant dangers [27]. A person’s resilience cannot be attributed to a single factor, but rather to the interactions between external and internal factors [28]. As a result, resilience is believed to be protective against psychiatric disorders.

The study discovered an inverse relationship between resilience and both burnout and stress in the healthcare professionals [29]. Researches indicated that a nurse’s individual resilience can assist in managing stress brought on by the COVID-19 pandemic [30, 31]. Research indicates that resilience can serve as a protective mechanism against the effects of burnout by mitigating workplace stress, enhancing personal resources, adapting to work challenges, and lessening stress’s adverse impacts [32]. The ability to withstand stress can forecast burnout in nurses, it is linked to fatigue and its influence on it [33]. Reports indicate that psychological resilience influences the extent of burnout, showing an inverse relationship with burnout caused by work and patients among emergency department staff [34, 35]. Meanwhile, nurses with high resilience endure high-pressure situations and recover more effectively from setbacks and maintain their well-being. This buffering effect reduces burnout by enhancing coping mechanisms and promoting a sense of control over one’s circumstances. Consequently, it was hypothesized that resilience would influence the connection between perceived stress and burnout.

Aim and hypotheses

Although it is well-known that nurses experience anxiety, stress and burnout, however, nurses’ anxiety and burnout and how anxiety contributes to burnout have not been studied in depth. Conducting this research is crucial to develop strategies for managing stress, preventing burnout, enhancing mental health outcomes, and boosting resilience among nurses working in infectious disease ward. Consequently, this research sought to (1) investigate the connection between anxiety, perceived stress and burnout among nurses working in infectious disease units and (2) examine this connection using a model of moderated mediation. Initially, we explored how perceived stress mediates the connection between anxiety and burnout. Then, we detected how resilience mediates the connection between perceived stress and burnout.

Materials and methods

Design, participants and procedure

The purpose of this cross-sectional study was to determine whether anxiety and burnout may be associated among infectious disease nurses in China. Cross-sectional research is typically faster and more cost-effective than longitudinal research since it gathers data at one specific moment. In our research, opting for a cross-sectional approach offered a momentary view of infectious disease nurses, enabling us to examine the rates of anxiety and burnout within this group. To mitigate potential bias in data collection, we used validated questionnaires and standardized procedures, and provide clear instructions to help participants recall information accurately.

Participants were selected based on a flowchart presented in Fig. 1. Through stratified cluster sampling, 18 out of 165 infectious hospitals (comprising 4 tertiary and 14 secondary public hospitals) and 32 out of 621 general hospitals (including 12 tertiary and 30 secondary public hospitals) were chosen from nine provinces between February to June 2023. In China, a hospital’s level correlates with the quality of its resources and standards. Clinical nurses in infectious disease ward were requested to fill out a survey including anxiety, burnout, psychological resilience and perceived stress, with the help of nurse managers. Eligible nurses were asked to join if they satisfied these conditions: (1) being a licensed nurse, (2) having over a year of experience in an infectious disease ward, (3) giving consent to participate, and (4) aged 18 ∼ 60 years old. Individuals were not included if they (1) held managerial positions, (2) declined to join, (3) had serious mental health conditions or (4) had been on leave for over six months prior to the survey. Participants received an optional, anonymous consent form clarifying that their employment and personal would remain unaffected by their decision to opt out. The Ethics Committee of the Chongqing University Cancer Hospital (CZLS2022022-A) sanctioned this research. This study was cross-sectional, so causality between burnout and anxiety or other variables could not be established. The data were adjusted for varying selection probabilities and response rates, and post-stratified to align the sample with the population distribution.

Fig. 1.

Flowchart of participants selection process

After reviewing the provided guidelines and agreeing to participate, individuals filled out an anonymous self-administered survey online. A sample of 1,618 nurses consented to join the study. Finally, we collected 1,603 questionnaires, of which 1,579 were valid for further analysis.

Measures

Demographics

The sociodemographic traits encompassed gender, work year, age, education, marital status, shift work status, weekly working hours, professional title and hospital classification.

Measurement of anxiety

The symptoms of generalized anxiety over the previous 14 days were assessed using the Generalized anxiety disorder-7 (GAD-7) elaborated by Spitzer [36]. The GAD-7 consists of seven questions, each rated on a four-point Likert scale from 0 (never) to 3 (nearly every day), with a total score between 0 and 21. Elevated scores reflect severe anxiety symptoms. Throughout China, the GAD-7 has been extensively used, and its reliability and validity have been well established [37].

Measurement of psychological resilience

Resilience can be assessed using the Conner-Davidson Resilience Scale (CD-RISC), developed by Connor and Davidson [38]. On the scale, 25 items are measured in three dimensions: strength, tenacity, and optimism. Ratings range from 0 (not at all) to 4 (extremely). Scores range from 0 to 100, where higher values signify greater resilience. Within the research, the Cronbach’s alpha coefficient was calculated as 0.950.

Measurement of perceived stress

Yang and Huang [39] updated the Chinese Perceived Stress Scale (CPSS) to estimate perceived stress levels experienced in the past month. The scale is composed of 14 items divided into 2 dimensions: sense of control (7 items) and sense of tension (7 items), each rated on a 5-point Likert scale ranging from 1 (“not at all”) to 5 (“always”). The overall score varied between 0 and 56. A coefficient of internal consistency of 0.82 was found for the overall scale, coefficient of 0.86, 0.77 for the two subscales, respectively [40]. If a participant had a total CPSS score ≥ 26, then this was judged as health risk pressure [41]. The Cronbach’s α was 0.859.

Measurement of burnout

Maslach Burnout Inventory-Human Service Survey (MBI-HSS) was frequently employed to assess burnout syndromes among nurses [42]. The MBI-HSS scale, consisting of 22-item, was divided into three subscales: EE (9 items), DP (5 items), and PA (8 items). For each item, burnout experiences and emotions were assessed using a 7-point Likert scale ranging from (“never”) to 6 (“daily”). Burnout showed a positive correlation with the total scores on the EE and DP subscales, while it was inversely related to PA scores. According to Maslach and colleagues, “low burnout” was characterized by EE of ≤ 16, DP of ≤ 6, and PA of ≥ 39; “moderate burnout” was identified EE scores between 17 and 26, DP scores between 7 and 12, and PA scores between 38 and 32; and “high burnout” was defined by EE scores ≥ 27, DP scores ≥ 13, and PA scores ≤ 31. The Cronbach’s α score values for the individual subscales varied between 0.768 and 0.882, while the overall scale had a Cronbach’s alpha coefficient of 0.780.

Date analysis

The analysis utilized SPSS (version 21.0) along with the Hayes SPSS macro program PROCESS to structure and evaluate the data, particularly focusing on the characteristics the moderated indirect effects. Initially, we performed both descriptive statistics and bivariate correlation analysis. To evaluate all regression coefficients, we applied bias-corrected percentile bootstrap method. Using 5000 bootstrap samples, 95% confidence intervals (CIs) were determined. There is a significant effect when the 95% CI does not include 0. A mediation model (Model 14) was chosen to analyze mediation effects in the current study.

Results

Sample characteristics

Descriptive analysis, independent sample t-test and one-way analysis of variance are presented in Table 1 to outline and contrast general demographic information, such as gender, work year, age, education level, marital status, shift work, weekly working hours, professional title, hospital level along with the distribution of scores for MBI-HSS dimensions, as shown in Table 1.

Table 1.

Participant characteristics (n = 1,579)

Variables	N(%)	Emotional exhaustion			Depersonalization			Reduced personal accomplishment
		M ± SD	t/F	P	M ± SD	t/F	P	M ± SD	t/F	P
Gender			1.459	0.147		-0.961	0.338		-2.423	0.015
Female	1464(92.72%)	19.96 ± 10.81			6.22 ± 5.33			27.61 ± 9.64
Male	115(7.28%)	18.35 ± 11.49			6.73 ± 5.53			29.87 ± 9.27
Age			3.138	0.024		7.027	<0.001		9.034	<0.001
≤ 25	320(20.27%)	21.06 ± 11.14			7.23 ± 5.93			27.65 ± 9.05
26 ~ 34	926(58.64%)	19.92 ± 10.81			6.24 ± 5.25			26.98 ± 9.68
35 ~ 44	258(16.34%)	18.38 ± 10.54			5.51 ± 4.83			29.89 ± 9.37
≥ 45	75(4.75%)	18.84 ± 10.90			4.89 ± 4.81			30.90 ± 10.58
Professional title			3.145	0.043		4.438	0.012		12.208	<0.001
Junior	1183(74.92%)	20.21 ± 11.03			6.48 ± 5.44			27.22 ± 9.58
Intermiddle	339(21.47%)	18.54 ± 10.18			5.68 ± 5.02			28.86 ± 9.44
Senior	57(3.61%)	19.97 ± 10.81			5.04 ± 4.80			32.86 ± 9.58
Education			0.423	0.736		0.834	0.474		6.379	<0.001
Primary	21(1.33%)	19.05 ± 14.46			7.14 ± 4.89			27.81 ± 11.58
Middle	411(26.03%)	20.35 ± 10.71			6.55 ± 5.59			26.61 ± 9.49
College/University	1098(69.54%)	19.68 ± 10.88			6.12 ± 5.25			28.00 ± 9.67
Postgraduates	49(3.10%)	19.59 ± 9.96			6.29 ± 5.61			32.59 ± 6.98
Marital status			2.504	0.082		4.241	0.015		2.910	0.054
Unmarried	561(35.53%)	20.36 ± 11.15			6.68 ± 5.63			26.99 ± 9.43
Married	988(62.57%)	19.46 ± 10.66			5.97 ± 5.13			28.21 ± 9.74
Divorced/Widowed	30(1.90%)	22.97 ± 11.22			7.70 ± 6.08			28.40 ± 8.68
Weekly working hours			-4.619	<0.001		-1.399	0.162		-4.284	<0.001
Less than 40 h	849(53.77%)	18.67 ± 10.17			5.71 ± 4.85			27.46 ± 9.49
40 h and more	730(46.23%)	21.21 ± 11.47			6.88 ± 5.81			28.14 ± 9.77
Shift work			2.706	0.007		2.197	0.028		-1.519	0.129
Yes	1013(64.15%)	20.40 ± 10.83			6.48 ± 5.46			27.50 ± 9.43
No	566(35.85%)	18.86 ± 10.86			5.86 ± 5.11			28.27 ± 9.95
Years of experience			3.582	0.006		5.135	<0.001		11.579	<0.001
Less than 5 years	431(27.30%)	20.57 ± 11.01			6.63 ± 5.54			28.05 ± 9.24
5–10 years	697(44.14%)	20.13 ± 11.02			6.58 ± 5.53			26.29 ± 9.56
11–15 years	218(13.81%)	19.96 ± 10.44			5.80 ± 4.72			28.86 ± 9.55
16–20 years	132(8.36%)	16.83 ± 9.99			5.34 ± 4.79			29.61 ± 9.83
More than 20 years	101(6.40%)	18.48 ± 10.46			4.55 ± 4.25			32.13 ± 9.60
Hospital level			1.785	0.168		3.188	0.041		5.317	0.005
Primary	46(2.91%)	21.96 ± 13.16			6.22 ± 4.88			26.20 ± 10.50
Secondary	348(22.04%)	20.47 ± 11.58			6.89 ± 5.92			26.44 ± 9.73
Tertiary	1185(75.05%)	19.58 ± 10.54			6.07 ± 5.17			28.23 ± 9.53

A total of 1,464 females and 115 males participated in the study. With respect to professional designations, 74.92% (1,183/1,579) held a junior position, 21.47% (339/1,579) held an intermediate-level position, and merely 3.61% (57/1,579) held a senior-level position. Among all participants, 27.30% (431/1,579) have been employed in infectious diseases ward for under 5 years; 44.14% (697/1,579) for 5 ∼ 10 years; 13.81% (218/1,579) for 11 ∼ 15 years; 8.36% (132/1,579) for 16 ∼ 20 years; and 6.40% (101/1,579) for over 20 years. The average years of clinical nursing experience was 3.25 ± 1.23 years.

Descriptive analysis

According to Maslach’s criteria for burnout and its aspects outlined in the Method section, burnout syndrome is widespread. Most participants (981/1,579) reported high burnout, while 24.0% (379/1,579) reported a moderate degree of burnout. According to Tables 2, 55.9% (883/1,579) of the nurses working in infectious disease ward experienced moderate to high levels of emotional exhaustion (EE), 33.5% (529/1,579) showed moderate to high levels of depersonalization (DP), and 86.1% (1,360/1,579) had moderate to high levels of reduced personal accomplishment (PA). The GAD-7 scores showed 627 (41.71%) nurses were normal (0 ∼ 4 points), 782 (42.78%) were mild (5 ∼ 9 points), 126 (7.98%) were moderate (10 ∼ 14 points), and 44 (2.79%) were severe (15 ∼ 21 points), out of which 952 nurses (58.29%) suffered from different degrees of anxiety.

Table 2.

Scores of nurse burnout

Maslach burnout inventory subscales	Mean ± SD	Low	Moderate	High
		N(%)	N(%)	N(%)
Emotional exhaustion(0–54)	19.85 ± 10.86	696(44.08%)	480(30.40%)	403(25.52%)
Depersonalization(0–30)	6.25 ± 5.35	1050(66.50%)	318(20.14%)	211(13.36%)
Personal accomplishment(0–48)	27.78 ± 9.63	219(13.87%)	379(24.00%)	981(62.13%)

As show in the Table 1, the results showed that age (F = 3.138, P = 0.024), professional title (F = 3.145, P = 0.043), weekly working hours (t=-4.619, P < 0.001), shift work (t = 2.709, P = 0.007), and years of experience (F = 3.582, P = 0.006) had significantly impacted EE. Additionally, age (F = 7.027, P < 0.001), professional title (F = 4.438, P = 0.012), marital status (F = 4.241, P = 0.015), shift work (t = 2.197, P = 0.028), years of experience (F = 5.135, P < 0.001), hospital level (F = 3.188, P = 0.041) had significantly affected DP. Furthermore, gender (t=-2.423, P = 0.015), age (F = 9.034, P < 0.001), professional title (F = 12.208, P < 0.001), education (F = 6.379, P < 0.001), weekly working hours (t=-4.284, P < 0.001), years of experience (F = 11.579, P < 0.001), Hospital level (F = 5.371, P = 0.005) had a significant impact on PA. As showed in the Table 3, the total resilience score is 55.66 ± 14.39, aligning with a previous study [43]. The average total score of EE is 19.85 ± 10.86, which had no statistical differences from the previously reported score [44]. However, the mean scores of GAD-7 was 5.45 ± 3.86 in this study, which was higher than the previous study (4.77 ± 3.50, n = 442) among infectious disease nurses [11]. In contrast to Shen Y’s report (21.00, n = 643) [23], the mean score of perceived stress in this study (25.04, n = 1,579) was significantly higher. These variations can be linked to our study’s significantly larger sample size and the fact that it was carried out in the post-COVID-19 era.

Table 3.

Descriptive statistics and bivariate correlations of the major study variables (n = 1,579)

	M ± SD	Anxiety	PSS	Resilience	MBI_EE	MBI_PA	MBI_DP
Anxiety	5.45 ± 3.86	1
PSS	25.04 ± 7.13	0.632**	1
Resilience	55.66 ± 14.39	-0.438**	-0.650**	1
MBI_EE	19.85 ± 10.86	0.535**	0.493**	-0.361**	1
MBI_PA	27.78 ± 9.63	-0.197**	-0.383**	0.556**	-0.061*	1
MBI_DP	6.25 ± 5.35	0.351**	0.316**	-0.237**	0.615**	-0.089**	1

Note PSS, Perceived Stress Scale; MBI, Maslach Burnout Inventory-Human Service; EE, Emotional exhaustion; DP, Depersonalization; PA, Personal achievement. ^**P<0.01, ^*P<0.05

Correlation analysis

Bivariate correlations between the main variables were also presented in Table 3. The anxiety was significantly positively correlated with perceived stress (r = 0.632), EE (r = 0.535), DP (r = 0.351) and negatively associated with resilience (r = − 0.438) and PA (r = − 0.197) (all P < 0.001). There was a notable positive correlation between perceived stress and both EE (r = 0.493) and DP (r = 0.316), while it was inversely correlated with PA (r = − 0.383) and resilience (r = − 0.650), with all correlations being statistically significant (P < 0.001). Resilience showed a strong positive correlation with PA (r = 0.556) and was inversely associated with EE (r = − 0.361) and DP (r = − 0.237; all P < 0.001).

Mediation effect analysis

A model 4 analysis was used in SPSS macro PROCESS software to investigate perceived stress’s role in the relationship between anxiety and EE. Model 1 in Table 4 indicated that the anxiety was a significantly positive predictor on EE (β = 0.535, t = 25.120, and P < 0.001). Model 2 indicated that anxiety was a strong positive predictor of perceived stress (β = 0.632, t = 32.387, and P < 0.001), and in turn, perceived stress significantly positively influenced EE (β = 0.259, t = 9.700, and P < 0.001). Furthermore, as illustrated in Model 3, the direct effect of anxiety on EE remained significant (β = 0.371, t = 13.903, and P < 0.001) even after incorporating mediating variables. In addition, the total effect (effect = 1.506, SE = 0.060, P < 0.001) and the direct effect (effect = 1.045, SE = 0.075, P < 0.001) of anxiety on EE were found to be significant in Table 5. There was an indirect effect of 0.461 (95% CI: 0.363 ∼ 0.558), accounting for 30.61% of the total effect. Within the 95% CI, there were no zero-point estimates. Hence, there was a statistically significant indirect effect, suggesting that perceived stress partially mediated the connection between anxiety and EE.

Table 4.

Mediation analysis (n = 1,579)

Variables	Model 1				Model 2				Model 3
	(MBI-EE)				(PSS)				(MBI-EE)
	β	t	Bootstrap 95%CI		β	t	Bootstrap 95%CI		β	t	Bootstrap 95%CI
			LLCI	ULCI			LLCI	ULCI			LLCI	ULCI
Anxiety	0.535	25.120	1.388	1.624	0.632	32.387	1.098	1.239	0.371	13.903	0.898	1.193
PSS									0.259	9.700	0.315	0.474
R 2	0.286				0.399				0.326
F	631.025***				1048.945***				381.177***

Note PSS, Perceived Stress Scale; EE, Emotional exhaustion; DP, Depersonalization. ^***P<0.001

Table 5.

Direct and indirect effects of the anxiety on emotional exhaustion

	Model path	Effect/Beta	SE	P	95% CI	Ratio of effect values
Total effect	Anxiety-Emotional exhaustion	1.506	0.060	0.0000**	[1.388,1.624]
Direct effect	Anxiety-Emotional exhaustion	1.045	0.075	0.0000**	[0.898,1.193]	69.39%
Mediating effect	Anxiety-perceived stress-Emotional exhaustion	0.461	0.049	0.0000**	[0.363,0.558]	30.61%

Note CI, Confidence Interval

Moderated mediation effect analysis

Table 6 illustrates that perceived stress was positively associated with EE, even after accounting for demographic variables such as weekly working hours and shift work. Following the inclusion in step 3, the findings indicated a distinct variance in EE while controlling for perceived stress, weekly working hours and shift work. In step 4, introducing the interaction term perceived stress × resilience led to notable outcome concerning EE. This suggested that resilience reduced the impact of perceived stress on EE. Based on a simple slope analysis, perceived stress predicted EE in both low (B = 0.854, t = 16.586, and P < 0.001) and high (B = 0.498, t = 9.503, and P < 0.001) levels of resilience, with a more pronounced correlation when resilience was lower (refer to Fig. 2).

Table 6.

Hierarchical multiple regression analyses of anxiety and resilience on emotional exhaustion (n = 1,579)

Steps and independent variables	Emotional exhaustion
	β	Total R2	ΔR2
Step 1
Weekly working hours	0.122
Shift work	-0.030	0.030	0.023
Step 2
Perceived Stress	0.610	0.256	0.251
Step 3
Resilience	-0.145	0.260	0.254
Step 4
Perceived Stress × Resilience	-0.205	0.278	0.272

Fig. 2.

The moderating effect of resilience on the relation between perceived stress and emotional exhaustion

Discussion

Lately, numerous studies have concentrated on the stress and exhaustion faced by clinical nurses, especially during the outbreaks of the COVID-19 pandemic [8, 9, 34]. The happiness and burnout levels of nurses have become a topic of heightened concern [45]. Due to the high-stress environments, extended shifts, and emotional difficulties that infectious disease nurses frequently encounter, the purpose of this study was to explore the relationship between anxiety, perceived stress, and burnout, revealing anxiety affects burnout directly as well as indirectly. Furthermore, this research introduces a novel moderated mediation model to elucidate the complex relationship between anxiety, perceived stress, and burnout in infectious disease settings. Drawing from earlier research and theoretical frameworks, we developed a model that incorporated the perceived stress as a mediator and resilience as a moderator in the relationship between anxiety and burnout (see Fig. 3).

Fig. 3.

The finalized moderated mediation model (n = 1,579). ***p<0.001

Previous research has underscored the significance of addressing burnout among nurses. Consistent with the findings of the present study, prolonged shift work schedules and excessive work-related stress emerged as primary contributors to burnout among nurses [46]. In our study, approximately four in five nurses presented burnout. This proportion was higher than reported previously [47, 48], particularly in terms of emotional burnout and diminished personal achievement, warranting greater focus. The perceived stress level in this study was notably higher than that of general nurses in China [49], highlighting the need to address perceived stress among infectious disease nurses. The average level of resilience in this study was consistent with previous study (55.88 ± 14.77, n = 845) [43]. Similarly, compared with the previously reported score for emergency department nurses (20.90 ± 11.78, n = 571) [44], the level of EE in our research had no statistical differences, which indicated that nurses working in emergency department and infectious disease department showed the similarly severe burnout.

A prior cross-sectional study conducted in China on nurse burnout and work quality revealed that approximately 64% of nurses experienced job burnout, with factors such as hospital classification, age, income, shift work assignments, and nurse-to-doctor ratio affecting quality of life and work [46]. Nevertheless, given that this prevalence significantly differs across various populations and clinical wards [50], it would be more suitable to analyze the contributing factors. This study showed that burnout is more pronounced in lower-titled and younger groups of nurses working in the infectious disease ward. The reason for this may be that nurses with lower-titled or young age are given simpler tasks and less responsibility; as a result, they may feel unfulfilled at work, leading to a lack of personal fulfillment [51]. In line with previous studies, our study showed that weekly working hours, shift work and years of experience were influencing factors in the emotional exhaustion dimension in the infectious disease ward [52, 53]. There are many work stressors for nurses in China, including heavy workloads and many non-nursing tasks. Nurses in the infectious disease unit encounter unique difficulties, such as frequent job-related exposure and heightened stress and exhaustion. However, 64.15% nurses in the infectious disease ward were involved in shift work and 46.23% nurses worked more than 40 h a week in our study. Obviously, nurses who were involved in shift work or worked longer hours per week are prone to exhaustion dimension, the nurse scheduling system should be optimized.

Furthermore, earlier studies demonstrated that 41.18% had different degrees of anxiety among healthcare professionals during the pandemic of COVID-19 [11]. The study found anxiety and perceived stress were positive predictors of Emotional Exhaustion (EE). These results align with earlier research indicating that increased stress and anxiety lead to more pronounced EE [54]. EE, considered the core component of the Maslach burnout model, is characterized by feelings of emotional depletion and fatigue [24]. Additionally, EE has been linked to other burnout symptoms such as depersonalization (DP) and cynicism [55]. The Hans Selye’s General Adaptation Syndrome (GAS) stress model has been proposed as a suggested as a basis for comprehending and forecasting the effects of EE. According to this model, stress responses progress through three stages: alarm, resistance, and exhaustion, with EE being a common outcome when stress reaches its peak. The combination of increased workplace stress and longer shifts may contribute to sleep deprivation and emotional exhaustion, which are major contributors to mental health problems and burnout [56].

The primary outcome of the analysis indicates that anxiety exerts both a direct and indirect impact on emotional exhaustion (EE), the latter mediated by perceived stress. These results were consistent with other research which found anxiety and perceived stress are closely related to job burnout [22]. Consistent with the previous hierarchical regression analysis [57], we found stress-burnout relationships were influenced by resilience. Among nurses in infectious disease ward with lower resilience, perceived stress has a more pronounced effect on EE compared to those with higher resilience levels. This study’s findings corroborate previous research on the protective effects of resilience in emotional exhaustion [46].

The ability of people to successfully handle and bounce back from difficult situations, known as resilience, has been associated with improved performance on the job and relieved emotional exhaustion (EE) [58]. Studies have repeatedly shown resilience was important to prevent mental health problems [59]. The resilience of nurses can be influenced by challenging work conditions, feelings of psychological emptiness, a disruption of inner balance, and mental conflict. Strategies for enhancing resilience encompass cognitive reframing, developing emotional fortitude, and fostering a sense of grounding connections, work-life balance, and reconciliation [60].

In the aftermath of COVID-19, nurses’ resilience is crucial due to various reasons. To begin with, nurses frequently encounter numerous pressures and obstacles in both their work and personal lives, which restrict their chances for relaxation and maintaining healthy social interactions. Developing high levels of psychological resilience can enhance their ability to communicate effectively and foster positive psychosocial relationships. Secondly, the prolonged shift work and high levels of work-related stress experienced by clinical nurses can adversely affect their quality of sleep. Neurobiological aspects of sleep quality and psychological resilience suggests that individuals with higher levels of resilience are less affected by perceived stress compared to those with lower resilience [61, 62]. Furthermore, people with strong resilience demonstrate qualities like effective coping skills, positive outlooks, realistic optimism, and mental adaptability, which can aid in working well with parents and teams in stressful scenarios [27]. The influence of resilience on anxiety and burnout revealed that nurses experiencing higher anxiety and lower resilience consequently faced intense burnout.

The implications of this study are both theoretical and practical. The findings may provide insight into the mechanisms underlying the relationships between anxiety, perceived stress, resilience, and burnout. Additionally, by highlighting the mediating role of resilience, this study may inform interventions aimed at implementing resilience training to alleviate burnout among nurses. Specifically, resilience training, drawing from positive psychology, cognitive behavioral therapy, and mindfulness, is anticipated to mitigate or eradicate burnout and anxiety in individuals experiencing high levels of occupational stress [63]. Resilience training is anticipated to mitigate burnout and anxiety in individuals experiencing high levels of occupational stress, particularly clinical nurses. Previous studies have shown that mindfulness-based resilience training (MBRT) effectively reduces stress and emotional burnout among healthcare professionals [64]. Therefore, further research is necessary to assess how resilience training can reduce anxiety-related burnout and improve job satisfaction among clinical nurses working in infectious disease ward. Thus, upcoming research could consider utilizing a longitudinal or an experimental research design.

Limitations

Alongside its merits, it is imperative to recognize certain limitations inherent to this study. This investigation utilizes a cross-sectional methodology, which inherently lacks the capacity to delineate temporal sequences, thereby complicating the establishment of causality. Absent knowledge regarding the chronological precedence of variables, we are precluded from confidently inferring causal relationships. Given that the cross-sectional framework provides solely a snapshot of the interrelations among burnout, anxiety, and perceived stress, it is incapable of depicting the evolution or transformation of these associations over time. Previous research on mediation analysis often overlooked temporal considerations, highlighting the potential for bias when employing a cross-sectional approach to investigate mediational dynamics. Drs. Judd and Kenny underscored the importance of incorporating prior assessments of both the mediator and the outcome within longitudinal mediation studies to minimize bias. Longitudinal investigations are indispensable for determining the causal influence of anxiety on burnout within nursing populations [65]. Moreover, this study’s reliance on self-reported measures may engender various biases. When addressing sensitive subjects, respondents may opt for answers deemed socially desirable instead of providing truthful responses. There is a likelihood that participants may exhibit systematic responding (such as selecting median options) or be swayed by question phrasing. Additionally, self-report data hinge on individuals’ subjective perceptions and interpretations, which might diverge from objective realities. Subsequent investigations should embrace a diversity of methodological approaches for evaluating stress and burnout.

Nursing implications

Our findings revealed that age, professional designation, weekly working hours, shift work, and years of working were statistically significant predictors of emotional exhaustion among nurses in infectious disease wards. Although causality remains unestablished, our results suggest that policymakers and nursing leaders should prioritize improvements in the nurse scheduling system, maintain equitable patient-to-nurse ratios, and foster professional development to alleviate or prevent job burnout in infectious disease nurses. Moreover, it is imperative for supervisors to encourage infectious disease nurses to incorporate established stress-reduction practices, such as mindfulness exercises, relaxation strategies, and efficient time management techniques. Promote the sharing of experiences, challenges, and coping mechanisms among nurses through peer support networks. To enhance resilience among infectious disease nurses, managers ought to periodically evaluate current resilience levels utilizing validated instruments such as the Resilience Scale, and design a customized resilience program comprising modules on stress management, emotional control, and positive mindset cultivation.

Conclusion

Nurses specializing in infectious diseases report elevated levels of anxiety, perceived stress, and burnout. Anxiety not only has a direct impact on burnout but also indirectly influences it through its effect on perceived stress, which is acknowledged as mediating the relationship between anxiety and burnout. This research highlights that resilience significantly contributes to stress management and the prevention of burnout. Further studies are warranted to elucidate the determinants that foster resilience, mitigate burnout, and enhance well-being. Additional research is imperative to uncover the causal mechanisms elucidating factors that enhance resilience, diminish burnout, and augment well-being. Employing longitudinal methodologies, researchers could examine temporal variations in burnout among nurses in infectious disease specialties, track fluctuations in anxiety and stress, and ascertain the impact of resilience training and stress management techniques on occupational satisfaction. The findings herein recommend that administrative bodies implement resilience training programs focusing on stress management, emotional regulation, and coping strategies. Supervisors are advised to devise targeted strategies promoting work-life equilibrium via policies enabling flexible scheduling and adequate leave, alongside establishing robust support systems encompassing mentorship initiatives, peer assistance groups, and access to counseling resources.

Abbreviations

EE

Emotional exhaustion

DP

Depersonalization

PA

Personal accomplishment

GAD-7

Generalized Anxiety Disorder-7

CD-RISC

Conner-Davidson Resilience Scale

CPSS

Chinese Perceived Stress Scale

MBI-HSS

Maslach Burnout Inventory-Human Service Survey

Author contributions

Conceived and designed the research: Ya-Lan Huang, Meng-Ting Chen, Li Liu. Performed the research: Ya-Lan Huang, Zong-Hua Wang, Yong-Guang Li, Zhi-Han Zhao, Wei-Yi Wang, Chang-Xia Cai, Xiu-Shuang Wu. Data analysis: Ya-Lan Huang, Zong-Hua Wang, Yong-Guang Li, Meng-Ting Chen; Wrote the paper: Ya-Lan Huang, Zong-Hua Wang, Meng-Ting Chen, Li Liu. The final manuscript was reviewed and approved by all the authors.

Funding

This research was supported by National Natural Science Foundation of China (No. 82404265), Natural Science Foundation of Chongqing, China (cstc2021jcyj-msxmX0400), Project for Performance Incentive Guidance of scientific research Institutions in Chongqing (cstc2022jxjl120007), Science and Technology Research Program of Chongqing Municipal Education Commission (Grant No. KJQN202300120), Chongqing medicinal biotech association of scientific research projects(cmba2022kyym-zkxmQ0011), and Key Research and Development Programme of Shanxi Province (2023-YBGY-147).

Data availability

Upon reasonable request, de-identified textual data can be provided to the corresponding author.

Declarations

Ethical approval

The Ethics Committee of Chongqing University Cancer Hospital (CZLS2022022-A) granted approval for this research, which adhered to the Declaration of Helsinki and the Good Clinical Practice standards set by the International Conference on Harmonization. Before the survey began, the study’s objective was clarified to all participants, and their informed consent was secured.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.