Work Stress, Risk Perception, and Social Support: Structural Equation Modeling of Healthcare Staffs’ Resilience

Abstract

The research aims to explore the current situation of resilience building of hospital medical personnel and its influencing factors, and try to put forward countermeasures and suggestions to improve the resilience of medical personnel, to alleviate the pressure of medical personnel’s work in normal times, and also to prepare for the effective enhancement of emergency treatment capacity in wartime. This study employed a cross-sectional design to survey medical staff from 2 hospital categories (infectious disease specialty hospitals and general hospitals). Using a stratified proportional sampling method, data were collected from 665 participants via scanned Wenjuanxing QR codes between January and March 2024. Data were analyzed using generalized linear models and structural equation modeling (SEM) to examine the relationships between resilience and its determinants. Significant differences in resilience levels were observed based on job position, title, income, and perceived occupational risk (P < .05). Social support emerged as a protective factor (β = .104, P < .001), while risk perception (β = −.135, P = .001) and work stress (β = −.195, P < .001) negatively impacted resilience. Individual factors, such as age and job tenure, moderated these effects. The work stress of medical staff in municipal hospitals is close to the imbalance threshold, and the overall level of mental toughness is good, but attention needs to be paid to their mental health. Work stress, social support and risk perception play an independent role in influencing psychological resilience, with social support playing the highest role in the pathway relationship. We should take comprehensive measures from the welfare distribution system, supportive work environment, regular health promotion, psychological intervention management, and social care system to promote timely relief of work stress and effective mental health interventions for medical staff in municipal hospitals.

What do we already know about this topic?

During the covid-19 epidemic, the hospital’s medical staff bore a large psychological burden. The concept of resilience is widely used by scholars to evaluate the mental toughness of individuals in the face of crisis.

How does your research contribute to the field?

Social support emerged as a protective factor, while risk perceptionand work stress negatively impacted resilience.

What are your research’s implications towards theory, practice, or policy?

Interventions intended to improve the resilience of medical staff should be based on the three dimensions of social support, risk perception, and work stress.

Introduction

Despite the continuous development and progress of human society, the situation of catastrophic events has become increasingly serious. ¹ Public health emergencies have become an important issue affecting national security and social stability. As the mainstay of healthcare service provision, medical personnel are at the front line of treating the sick and saving lives, and their resilience in the face of public health events is related to the resilience level of the healthcare service system. As frontline responders, medical staff face heightened risks of infection, prolonged work hours, and emotional distress, which can significantly impact their mental health and resilience. Despite growing recognition of these challenges, there remains a critical need to understand the factors that influence resilience among healthcare workers, particularly in high-stress environments like municipal hospitals.

Research has shown that public health emergency have resulted in unprecedented psychological impact on healthcare workers. ² The occurrences have led to a sudden increase in their workload, making their tasks more demanding and complex. Healthcare workers constantly face psychological burdens such as concerns and fears of infection risks, the urgency of epidemic prevention and control, among others. ³ The World Health Organization (WHO) issued a warning after the COVID-19 pandemic, stating that public health emergencies may have negative effects on the mental well-being of healthcare workers. ⁴ Evidence from the spread of infectious diseases also indicates that healthcare workers are more likely to face risks of short-term or long-term mental health issues. ⁵ During the global COVID-19 pandemic, the high rates of anxiety and depression among healthcare workers ranged from 23.2% to 40.0% and 22.8% to 37.0%, respectively. ⁶ Up to one-third of frontline healthcare workers are experiencing significant psychological distress, which can even lead to harm to their individual health, impacting their ability to function normally. ⁴

As far as Beijing is concerned, municipal hospitals, as the main medical bodies for emergency treatment of public health emergencies, have assumed most of the medical treatment, and other responsibilities. Since the COVID-19 pandemic, municipal hospitals have admitted and treated a total of 6090 confirmed patients, accounting for 97.3% of the city’s total, and 132 severely ill and critically ill patients, accounting for 93.6% of the city’s total. Therefore, the mental health of the medical staff of Beijing municipal hospitals is particularly important in the event of a public health emergency.

Mentally healthy healthcare workers often possess the ability to overcome adversity in the workplace. ⁷ They are better able to adjust, balance, and control themselves in unfavorable circumstances and find solutions to challenges. ⁸ Collectively, these abilities are known as resilience. Research has shown that resilience can positively influence employee satisfaction and creativity, which in turn enhances an individual’s psychological state when faced with unexpected changes and crises. ⁹ Highly resilient healthcare workers are more calm and organized in the face of crisis, are less likely to be overwhelmed by stress and maintain job sustainability. ¹⁰

Research has found that resilience is influenced by a variety of factors. Among them, acquired and modifiable factors such as workplace stress, social support and risk perception are important determinants of resilience.. ¹¹ The workplace stress is a reflection of the individual’s work effort. Researchers have found that the longer an employee works, the lower his or her level of stress resilience, with employees who work more than 10 h a day having significantly lower levels of stress resilience than those who work 8 h a day. ¹² Medical staff working in high-pressure work environments, such as medical intensive care units, for long periods of time reduced their stress tolerance levels. ¹³

Social support is an essential social resource for individuals. High levels of social support can promote resilience and positive psychological outcomes to sustain an individual’s physical and mental health. ¹⁴ Studies have found that social interactions among colleagues in hospitals, such as mutual understanding and bonding, can increase resilience levels among medical staff. ¹⁵ At the same time, institutionalized protection by organizations or superiors plays an equally irreplaceable role. A study of Chinese anti-epidemic medical personnel found that protection training and psychological counseling services provided by hospitals were protective factors for anxiety. ¹⁶ Organizational support to reduce the risk of secondary traumatic stress through increased resource accessibility. ¹⁷

Risk perception is an individual’s subjective feelings and perceptions of external objective risks. Numerous studies have confirmed that the risk of occupational exposure of medical staff under public health events is an important stressor for them, which induces mental health problems in medical staff. ¹⁸ Risk perception plays a negative role in employee resilience. The higher the risk perception of employees, the more likely they are to develop negative emotions such as anxiety, which can also have a negative impact on the collective, which in turn affects job resilience. ¹⁹ Individual factors such as gender, ²⁰ education, ²¹ age ²² and number of years working in the current hospital ²¹ have been widely used as moderating variables in the study, and differences between individuals vary in the role of work, social support, and risk perception on the level of resilience.

In most of the studies, researchers have examined a single influencing factor for resilience. And there is no consistent conclusion about whether medical staff resilience is related to individual and external factors. ²³ Literature studying the factors influencing the resilience competence of medical personnel based on the perspective of resilience competence enhancement of medical personnel is relatively small. No systematic and comprehensive research has been conducted. ²⁴ The content of the construction of the mental toughness structure is in a state of ambiguity. ²⁵ Based on the above discussion, this study constructs a model of resilience level of medical staff based on the 2-factor theory ²⁶ and DeSeCo Australian model, ²⁵ including risk perception, work stress, social support, and incorporates individual factors as moderating variables. The theoretical model is shown in Figure 1, and the hypotheses are proposed:

Figure 1.

Theoretical model diagram.

1. Risk perception has a negative impact on the resilience of medical staff.

2. Work stress has a negative impact on the resilience of medical staff.

3. Social support has a positive impact on the resilience of medical staff.

4. Individual factors play a moderating role in the relationship between work stress, social support, risk perception, and the resilience of medical staff.

Methods and Data

Study Design and Participants

This study employed a stratified proportional sampling approach to select participants from Beijing municipal hospitals. Among the 22 municipal hospitals in Beijing, we categorized them into 2 groups: infectious disease specialty hospitals (n = 4) and general hospitals (n = 18). Since hospitals specializing in infectious diseases are the mainstay of care under public health events, general hospitals play a relatively equal role under such events. Therefore, this study was stratified by infectious disease specialty hospitals and general hospitals. One hospital was randomly selected from each category, and 10% of the medical staff from each hospital were sampled, ensuring representation across roles (doctors, nurses, medical technicians, etc.). ²⁷

Medical staff of municipal hospitals in Beijing, mainly including doctors, nurses, medical technicians, and administrators, etc. The inclusion criteria were: (1) those who voluntarily participated in the survey after informed consent; (2) those who were formally on duty; and (3) those who had experienced COVID-19 pandemic and had been working in the organization for more than 3 months prior to the event. The exclusion criteria were (1) medical staff in rotation, internship, or in the training stage; (2) those who had not undergone standardized training; and (3) those who were not on duty.

Data Collection

Data were collected from January to March 2024. Participants completed the relevant scales via mobile devices by scanning Wenjuanxing QR codes, yielding a final sample of 665 respondents.340 medical personnel were selected from infectious disease specialized hospitals and 325 medical personnel were selected from general hospitals to participate in the questionnaire response, and the number of study subjects met the proportion requirement of the number of questions in the scale. It excluded individuals who lacked clear information on personal characteristics or resilience levels during the survey process. After extreme value cleaning of the survey data, a total of 620 healthcare professionals were ultimately included in this research. The response rate for this survey was 93.2% which is considered a reasonable response rate. See Figure 2.

Figure 2.

Flowchart for inclusion of study participants.

The questionnaires were uniformly distributed and collected by the medical department in the administrative department, by the head nurse of each department in the clinical department, and by the chief of each department in the medical-technical department. Electronic questionnaires were used for collection and reverse scoring questions were designed in the questionnaires.

We confirm that all methods were carried out in accordance with relevant guidelines and regulations. This study was approved by the Medical Ethics Committee of Capital Medical University (Z2024SY035) on 2024.7.17. Because the data of this study were collected using a web-based survey, the data were de-identified, the subjects’ privacy was not disclosed, and the exemption of informed consent would not adversely affect the rights and health of the subjects, an application for exemption of informed consent was granted by the Ethics Committee.

Measurement of the Variables

The content of the survey includes 5 components: an assessment of healthcare professionals’ resilience levels, risk perception evaluation, social support assessment, environmental stress evaluation, and socio-demographic information. The questionnaire used in this study comprises 5 sections: a self-constructed demographic questionnaire, an perception of risk in epidemic and pandemic prevention and control efforts scale, a brief effort-reward imbalance questionnaire (ERI), a social support rating scale (SSRS), and a connor-davidson resilience scale (CD-RISC). See Table 1.

Table 1.

Variables and Assignment Status.

Observed variables	Definition	Assignment	Scale	Cronbach’s α
Severity	Infection with the COVID-19 virus can cause significant harm to physical health.	1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Always	Perception of risk in epidemic prevention and control efforts	.912
Susceptibility	Everyone is more susceptible to infections.
Controllability	Controlling the COVID-19 pandemic proves to be quite challenging.
External contributions	The observable quantitative and qualitative workload, along with the recent increase in workload, includes factors such as time, mental effort, physical exertion, and responsibilities.	1 = Totally disagree, 2 = Disagree, 3 = Agree, 4 = Totally agree	ERI	.777
Work compensation	Including financial resources, respect (both acknowledgment and recognition), and professional opportunities.
Intrinsic investment	The unobservable increase in workload exceeds the normal capacity of the individual.
Objective support	The actual support received by healthcare professionals.	The reverse scoring method at Level 4 indicates that a higher score reflects a greater level of social support.	SSRS	.686
Subjective support	Emotional support that healthcare professionals can experience.
Utilization of support	Reflect the individual’s proactive engagement with various forms of social support, including methods of expression, avenues for seeking assistance, and participation in activities.
Tenacity	Maintaining composure, exhibiting determination, responding swiftly, and possessing a sense of control are essential when confronting challenges.	1 = Never, 2 = Rarely, 3 = Sometimes, 4 = Often, 5 = Always	CD-RISC	.972
Self-empowerment	Experiencing setbacks not only allows for recovery but also fosters development and growth.
Optimism	Having confidence in overcoming adversity and viewing situations from a positive perspective.

Note. Manifest Variable or Observed Variable is a variable that can be directly observed. Latent Variable is usually a variable that cannot be directly observed and needs to be estimated with the help of an exogenous measure. ³²

• (1) Socio-Demographic Information

This includes information such as gender, age, residency status, education level, job title, employment type, years of service, average monthly income, and position.

• (2) Perception of Risk in Epidemic and Pandemic Prevention and Control Efforts Scale

The COVID-19 risk perception scale ²⁸ compiled by Cui et al Previous studies have shown that this scale has a good reliability and validity. The scale consists of 9 questions and 3 dimensions, namely controllability (it is difficult to control COVID-19), severity (novel coronavirus causes serious damage to physical health) and susceptibility (everyone is susceptible to infection). The scale applies the Likert 5-point scoring criterion, and higher scores of individuals represent a higher level of risk perception. The Cronbach’s α coefficient of this scale in this study was 0.912, indicating excellent internal consistency reliability.

• (3) Effort-Reward Imbalance Scale (ERI)

This was later revised and translated into Chinese by Dong and Shi. ² The questionnaire measured 3 aspects with 15 core questions, including 3 questions of external effort, 4 for internal input, and 8 work returns. The 4-point score is used, and the formula of “pay dimension score/return dimension score×0.375” is used to express the imbalance level of employee pay-return. When the ratio of pay/return is greater than 1, the epidemic prevention and control work of employees is greater than the return, indicating that they are in a bad situation. Otherwise, it indicates that its working conditions are better. The Cronbach’s α coefficient of this scale in this study was 0.777, indicating good internal consistency reliability.

• (4) Social Support Rating Scale (SSRS)

The Social Support Rating Scale (SSRS) developed by Xiao Shuiyuan was used, ²⁹ which contains objective support dimensions (3 entries), subjective support dimensions (4 entries), and utilization of support dimensions (3 entries). The scale is scored on a 4-point reverse scale, with higher scores indicating higher levels of social support.

• (5) Connor-Davidson Resilience Scale (CD-RISC)

Developed by Connor KM, ³⁰ Chinese scholars Xiaonan and Jianxin ³¹ adapted this scale for Chinese employees. This scale includes 25 items categorized into 3 dimensions: resilience (13 items), self-enhancement (8 items), and optimism (4 items). A higher score indicates a greater level of psychological resilience. In this study, the Cronbach’s α coefficient for the scale was 0.972, demonstrating excellent internal consistency reliability.

Statistical Analysis

Rank-sum tests were employed for categorical variables, using risk perception factors, workplace stress factors, social support factors, and individual factors as independent variables, with resilience and dimension scores as dependent variables (P < .05). A generalized linear model analysis explored the impact of risk perception, workplace stress, social support, and individual factors on the resilience levels of healthcare staff in top-tier hospitals (P < .05). Mplus 8.3 software was used to construct a structural equation model based on the research context and survey data, conducting confirmatory factor analysis with parameter estimation via weighted least squares mean and variance (WLSMV). Latent variables were defined through fixed loadings, employing 2-tailed testing. ³² All statistical analyses were conducted at α = .05, with P < .05 indicating significant differences.

Statement

The study has followed the relevant EQUATOR guideline in the Methods section. ³³

Results

Variability Analysis of Resilience Levels of Medical Personnel With Different Characteristics

A total of 620 medical staff of municipal hospitals were included in this study, of which 74.7% were female and 25.3% were male. The age distribution was mainly centered on 31 to 50 years old (71.2%), with 18.1% and 10.8% being 30 years old and below and 50 years old and above, respectively. Educational attainment was dominated by bachelor’s degree (50.5%), followed by master’s degree (18.9%) and doctoral degree (20.5%). Marital status was predominantly married (74.4%), with 21.3% unmarried. The highest percentage of job titles was nursing (42.3%), followed by medical (27.6%). In terms of job titles, junior and intermediate job titles accounted for a higher percentage, 37.3% and 36.1% respectively. The percentage of staff in the establishment was 80.0%. Working hours were mainly ≤ 8 h per day (55.0%), with 8 to 10 h accounting for 37.4%. Individual monthly income was mainly concentrated in 8000 to 15 000 yuan (67.1%), with ≤8000 yuan and >15 000 yuan accounting for 12.1% and 20.8%, respectively. Medical staff who perceived the risk of occupational exposure accounted for 65.3%. In COVID-19 pandemic, work stress manifested as giving more than rewarding accounted for 40.8%, and overloaded state accounted for 4.5%.

Using non-parametric tests, we analyzed the resilience levels of healthcare personnel across various characteristics. For binary variables, we applied the Mann-Whitney U test. For multi-category variables, we used the Kruskal-Wallis H test. The results, as indicated in Table 2, show that differences in positions, job roles, monthly household income, return on investment, excessive commitment, and perceived occupational exposure risk have statistically significant impacts on resilience levels (P < .05).

Table 2.

Analysis of Resilience Scores and Variability in Resilience Ratings Across Different Feature Sample Populations [n (%), n = 620].

Factor	Frequency	Rank average	Z/H
Group			46 290.500
Hospital A	295 (47.6)	304.93
Hospital B	325 (52.4)	315.57
Position			17.058***
Department head	12 (1.9)	368.79
Deputy head	19 (3.1)	309.74
Party branch secretary/deputy secretary	2 (0.3)	230.75
Head nurse	89 (14.4)	369.49
Team leader/medical team leader	36 (5.8)	352.93
General medical staff	442 (71.3)	295.26
Others	20 (3.2)	282.18
Gender			33 987.500
Male	157 (25.3)	325.52
Female	463 (74.7)	305.41
Age (years)			0.119
≤30	112 (18.1)	313.21
30-40	262 (42.3)	312.11
40-50	179 (28.9)	308.05
50-60	67 (10.8)	306.22
Marital status			1.817
Single	132 (21.3)	306.60
Married	461 (74.4)	311.85
Divorced	25 (4.0)	319.28
Widowed	2 (0.3)	147.00
Educational attainment			2.104
Doctorate	127 (20.5)	314.93
Master’s degree	117 (18.9)	290.73
Bachelor’s degree	313 (50.5)	317.59
Associate degree and below	63 (10.2)	303.06
Professional titles			2.986
Senior professional	45 (7.3)	341.64
Associate professional	93 (15.0)	320.25
Intermediate professional	224 (36.1)	310.30
Entry-level professional	231 (37.3)	304.98
None	27 (4.4)	273.89
Years of service			6.158
≤5	130 (21.0)	310.03
6-10	90 (14.5)	329.83
11-15	145 (23.4)	299.64
16-20	68 (11.0)	332.18
21-25	63 (10.2)	274.42
26-30	66 (10.6)	301.83
≥31	58 (9.4)	332.34
Employment methods			27 960.500
In-position	496 (80.0)	304.87
Non-position	124 (20.0)	333.01
Job positions:			12.851**
Healthcare	171 (27.6)	296.64
Nursing	262 (42.3)	337.84
Medical Technology	89 (14.4)	300.33
Pharmaceutical Department	46 (7.4)	256.66
Logistics	17 (2.7)	292.82
Administrative Management	35 (5.6)	278.74
Personal monthly income levels (Yuan)			5.132*
8000 and below	75 (12.1)	274.85
8001~	416 (67.1)	309.76
15 001~	129 (20.8)	333.62
Household monthly income level (Yuan)			9.302**
8000 and below	92 (14.8)	285.79
8001~	250 (40.3)	292.69
15 001~	278 (44.8)	334.69
Return on investment			30 763.500***
≤1	367 (59.2)	353.18
>1	253 (40.8)	248.59
Overcommitment			5346.000***
Low load	592 (95.5)	315.47
Overload	28 (4.5)	205.43
Daily working hours (h)			0.101
≤8	341 (55.0)	311.28
8-10	232 (37.4)	310.36
10-12	39 (6.3)	308.33
>12	8 (1.3)	291.63
Perception of occupational exposure risks			32 369.500***
Yes	405 (65.3)	282.92
No	215 (34.7)	362.44

^*P < .1. **P < .05. ***P < .01.

Analysis of Factors Influencing the Level of Resilience of Medical Personnel

A generalized linear model analysis using resilience levels of healthcare personnel as the dependent variable reveals several influencing factors. These factors include gender, marital status, years of service, input-output ratio, total social support score, and risk perception score. The results indicate that male healthcare workers exhibit higher resilience than their female counterparts. Additionally, married healthcare personnel demonstrate greater resilience compared to those who are single. Healthcare workers with 21 to 25 years of experience possess higher resilience levels. Those who perceive that they give more than they receive exhibit higher levels of resilience. Furthermore, higher total social support scores correlate with greater resilience, while lower risk perception scores are associated with enhanced resilience levels, as detailed in Table 3.

Table 3.

Analysis of Factors Affecting the Resilience Levels of Healthcare Personnel.

Variable	Partial regression coefficient	S.E.	χ2	P
Intercept	45.586	33.8058	1.818	.178
Gender	3.998	1.4841	7.257	P < .001
Marital status (with unmarried as the reference category)
Married	4.858	1.9195	6.406	.011
Divorce	0.297	3.5068	0.007	.933
Widowhood	9.411	10.4058	0.818	.366
Years of service (with ≤5 as the reference)
6-10	−1.041	2.4486	0.181	.671
11-15	4.042	2.8706	1.983	.159
16-20	0.974	3.8085	0.065	.798
21-25	8.935	4.4554	4.022	.045
26-30	4.037	4.5595	0.784	.376
≥31	−1.160	5.6309	0.042	.837
Return on investment	−4.745	1.2996	13.329	P < .001
Total score of social support	1.019	0.0799	162.527	P < .001
Total score of risk perception	−0.229	0.0882	6.754	P < .001

An Analysis of the Role Pathways Affecting Resilience Levels in Medical Personnel

• (1) Structural Equation Modeling and Goodness-of-Fit Test

Structural equation modeling was constructed using risk perception, work stress, and social support as independent variables (X) and mental toughness as dependent variable (Y). The measurement model was first modified according to the modification index (MI) and the model fit was tested, and the questions in the scale were reasonably censored to make the model fit to the specification. Based on the survey data, the correlation matrix, and model fitting information, establish the variable set and structural equation model. The reliability, convergent validity, and discriminant validity analysis of the smoking cessation effect structural equation model are shown in Table 4. In this table, the standardized factor loadings for each dimension range from 0.4 to 1.0, with composite reliability greater than 0.6 and convergent validity (AVE) exceeding 0.4. The square root of AVE values is greater than the correlations with other related constructs, indicating the presence of discriminant validity. The model fit parameter reference values are presented in Table 5, where the structural equation model fit indices are: χ²/df = 3.141, CFI = 0.934, TFI = 0.929, RMSEA = 0.059, SRMR = 0.049. These suggest that the resilience level structural equation model exhibits a good fit.

Table 4.

Analysis of Reliability, Convergent Validity, and Discriminant Validity.

Dim	Items	Item reliability	Composite reliability	Convergence validity	Discriminate validity
		STD.LOADING	CR	AVE	FX	HJ	SH	RX	GT
Risk perception	6	0.633-0.842	0.883	0.560	0.748
Workplace stress	5	0.544-0.856	0.857	0.550	0.307	0.742
Social support	3	0.425-0.884	0.785	0.569	−0.155	−0.164	0.754
Resilience	18	0.699-0.866	0.972	0.659	−0.252	−0.297	0.424	0.812
Individual factor	3	0.637-0.994	0.887	0.730	−0.044	0.131	0.131	−0.013	0.855

Note. The bold values along the diagonal represent the square root of AVE, while the triangles below signify the Pearson correlations of the dimensions.

Table 5.

Indicators of Model Fit.

Fit indicators	Recommended value	Model metrics	Compliant
ML χ2	The smaller the better	1438.406
Df	The bigger the better	371
χ2/df	χ2/df<5	3.141	Compliant
CFI	>0.9	0.934	Compliant
TLI	>0.9	0.929	Compliant
RMSEA	<0.08	0.059	Compliant
SRMR	<0.08	0.049	Compliant

Note. RMSEA = Root Mean Square Error of Approximation; CFI = Comparative Fit Index; TLI = Tucker-Lewis Index; SRMR = Standardized Root Mean Square Residual.

• (1) Structural Equation Modeling Path Analysis

The structural equation modeling results for resilience levels indicate that risk perception is explained by the statements: “I believe that contracting COVID-19 could lead to severe long-term effects” (β = .634, P < .001), “I feel that my likelihood of infection is high” (β = .782, P < .001), “I think I am more susceptible to infection than others” (β = .805, P < .001), “I believe the outbreak in my area is severe, increasing my risk of infection” (β = .842, P < .001), “I think controlling the spread of this pandemic is quite challenging” (β = .769, P < .001), and “I feel that the pandemic is far from over, and the risk of infection remains” (β = .633, P < .001). Work stress is explained by: “Compared to before the pandemic, the enormous workload causes me constant time pressure” (β = .814, P < .001), “Compared to before the pandemic, I frequently face interruptions while working” (β = .759, P < .001), “Compared to before the pandemic, my work stress has significantly increased” (β = .856, P < .001), “Due to the pandemic, I find it easy to become irritated from work-related stress” (β = .697, P < .001), and “Those who know me well feel that I am sacrificing too much for my work due to the pandemic” (β = .544, P < .001).

Social support is explained by the “support and care received from family members” (β = .425, P < .001), the “sources of financial assistance and practical problem-solving help received during past emergencies” (β = .861, P < .001), and the “sources of comfort and concern provided during past emergencies” (β = .884, P < .001). Resilience is detailed by responses to the 18 items such as “I can adapt to change” (β = .775, P < .001), “I can cope with whatever happens” (β = .812, P < .001), “Past successes give me confidence to face challenges” (β = .855, P < .001), and “I can see the humorous side of things” (β = .820, P < .001).“Perceived risk” (β = −.135, P = .001), “work stress” (β = −.195, P < .001), and “social support” (β = .371, P < .001) have an impact on the level of resilience of medical staff. Hence Hypothesis 1, 2, and 3 are supported.

Individual factors are explained by “age” (β = .994, P < .001), “professional title” (β = .637, P < .001), and “work experience” (β = .893, P < .001). Individual factors were included in the model as moderating variables. The results are shown in Table 6. The interaction term between work stress and individual factors was a significant predictor of resilience (β = .104, P = .026), thus the moderating effect of individual factors between work stress and resilience was significant. The interaction term between risk perception and individual factors was a significant predictor of resilience (β = −.100, P = .032), thus the moderating role of individual factors between risk perception and resilience was significant. The interaction term between social support and individual factors was a significant predictor of resilience (β = .084, P = .279), so the moderating effect of individual factors between social support and resilience was not significant. Hence Hypothesis 4 is partially supported. (1) Risk perception: lower awareness of external risks enhances resilience levels among healthcare workers. (2) Work stress: reduced self-perception of input and overexertion boosts resilience levels. (3) Social support: higher levels of received support improve resilience. (4) Moderating variable: increased age, length of service, and professional title mitigate the negative impact of risk perception on resilience and elevate the adverse effects of work stress on resilience. Refer to Table 6 and Figure 3.

Table 6.

Analysis of Research Model Hypotheses.

DV	IV	Estimate	S.E.	Est./S.E.	P-value	R 2	Hypothesis
Resilience levels	Risk perception	−0.135	0.041	−3.322	.001	0.249	Support
	Workplace stress	−0.195	0.041	−4.784	***		Support
	Social support	0.371	0.037	9.954	***		Support
	Workplace stress×individual factors	0.104	0.047	2.232	.026		Support
	Risk perception×individual factors	−0.100	0.047	−2.149	.032		Support
	Social support×individual factors	0.084	0.077	1.083	.279		Not support

^***P < .001.

Figure 3.

Pathway of factors influencing the resilience level of healthcare personnel.

Note. ^aindicates P < .01; ^bindicates P < .05; the path coefficients in the figure represent standardized coefficients.

Discussion

Comprehensive Interventions Mitigate Occupational Risk Perception in Healthcare Workers

Based on the study’s objectives, the first hypothesis was developed related to the negative association of risk perception with resilience levels. The result of the study shows that there is a significant negative effect of risk perception on resilience levels. The result is consistent with previous studies. ³⁴ Risk perception is an important factor that depletes the psychological resources of healthcare workers, ³⁵ disrupting their cognition and attention and depleting their coping resources. ³⁶ Medical personnel are prone to negative psychological states such as emotional instability, self-doubt, and guilt. ³⁷ Several scholars have studied medical staff under COVID-19 pandemic and found that high levels of risk perception can undermine medical staff’s resilience levels and affect their quality of work life. ³⁸ Even medical staff may be triggered to be over-vigilant due to high risk perception. This in turn weakens emotional regulation and problem-solving ability, creating a vicious cycle of “high risk perception - low resilience.” ³⁹ Health workers serve as important human resources in the health system. Their poor performance also affects the quality and efficiency of the health system and the community as a whole. ⁴⁰

Developing and applying intelligent diagnostic and treatment assistants to alert patients to their condition with recommended levels of protection and clarify risk levels in order to reduce the level of risk anxiety among medical staff. Starting from enhancing the professionalism of medical staff, scenario simulation training, such as desensitization training in high-risk scenarios such as simulated aerosol exposure and needlestick injuries, has been carried out to strengthen the protective skills of medical staff. Transforming the healthcare environment, such as applying advanced technology to quickly build negative pressure wards, to reduce the risk of disease transmission and boost the confidence of medical staff.

Interventions Reduce Healthcare Workers Workload and Emotional Labor Stressors

Based on the study’s objectives, the second hypothesis was developed related to the negative association of workplace stress with resilience levels. The result of the study shows that there is a significant negative effect of workplace stress on resilience levels. The results of the study indicate that work stress has a significant negative effect on the level of resilience. That is, the higher the workload, the lower the resilience. This finding matches existing studies. Research shows medical professionals often develop psychological imbalance when over-committed. This imbalance subsequently reduces work efficiency and quality. ²

Significant increase in working hours and workload of medical staff during the COVID-19 pandemic period. ⁴¹ This may not only add to the mental burden of medical staff, but may also lead to anxiety and depression and physical discomfort. ⁴² Increased work stress inhibits medical staff from obtaining positive information, ⁴³ reduces their initiative and concentration at work, and seriously affects the quality of their work. ⁴⁴ On the contrary, appropriate rest and workload adjustment are conducive to reducing medical staff burnout and effectively improving work efficiency. ⁴⁵

Therefore, work hours and work stress during public health emergencies should be systematically coordinated. This necessitates optimizing workflow protocols for shift rotations, rest periods, and duty rotations while implementing dynamic workforce allocation adjustments to align staffing with fluctuating operational demands. Leveraging AI and internet-based information systems, flexible work modalities should be implemented to enable remote handling of contact-free tasks, thereby minimizing occupational exposure duration in high-risk environments. Concurrently, comprehensive paid leave policies during health emergencies must be established alongside prioritizing post-event psychological recovery support for healthcare workers, through coordinated efforts to holistically alleviate their occupational burden.

Enhancing Healthcare Workers Well-Being Through Social Support Systems

Based on the study’s objectives, the third hypothesis was developed related to the positive association of social support with resilience levels. The result of the study shows that there is a significant positive effect of social support on resilience levels. This is consistent with the results of several existing studies. ⁴⁶ Social support was found to be an important psychoprotective factor. ⁴⁷ Social support can buffer psychological damage under catastrophic events by reducing or equalizing stressful life events. ⁴⁸ At the same time, changing healthcare professionals’ perceptions of stressful events helps individuals in challenging situations and improves their coping skills. ⁴⁹ The higher an individual’s level of social support, the more he or she feels understanding and respect from friends, family, and coworkers, which helps them to engage in self-affirmation and thus become more optimistic. ⁴⁸ These supportive forces in adversity can enhance the ability of medical staff to resist stress and increase their level of resilience. ⁵⁰ Medical personnel are able to obtain social support by confiding their emotions with family and friends and communicating harmoniously with their colleagues. ⁵¹

Accordingly, activities to provide vacation care for the children of medical personnel are constantly being improved, and green channels for medical personnel’s families to seek medical care are being opened. This will increase the level of support for the families of medical personnel. To popularize the complexity of medical decision-making through the media, to reduce the public’s cognitive bias, and to enhance the public’s understanding of and support for the work of medical professionals. To build an online communication platform to popularize medical knowledge and expand the interaction channels between medical personnel and the society, so as to enhance the level of social support for medical personnel.

The Moderating Role of Individual Factors

The results of this study showed that individual factors consisting of age, length of service and title of medical staff significantly moderated the dynamic relationship between job stress, risk perception, and psychological resilience. Based on Super’s theoretical framework for career development, ⁵²  “occupational seniority” is often used as a combination of experience, skills, and job level, so the individual factor in this study can be used as a proxy for occupational seniority. The high fit of the measurement model of occupational seniority as a latent variable integrating age (career stage), length of service (accumulation of experience) and title (achievement level) validates the construct’s structural validity. This is highly compatible with the findings of foreign scholars who used age, length of service and job title alone as moderating variables.^53,54 The results indicate that the negative impact of job stress on mental toughness diminishes as medical professionals’ professional seniority rises, while the negative impact of risk perception on mental toughness rises. The reason for this may be that, with richer experience in responding to public health incidents, highly senior medical personnel are able to more effectively mitigate the psychological depletion caused by work stress. Secondly, as senior personnel are required to take on more complex risk assessment and decision-making responsibilities, they have increased sensitivity to potential risks and therefore may require more energy and dedication to accomplish their tasks.

Therefore, a tiered approach to resilience enhancement measures should be implemented for medical staff. For newly recruited medical staff or those who have not worked for a long time, their horizons and ability to deal with public health emergencies should be popularized and broadened; for more senior medical staff and cadre leaders, the relevant departments should introduce a more detailed code of practice for dealing with public health emergencies, so as to reduce the scale and number of problems that may occur.

Limitation

This study has a number of limitations. First, it was not possible to establish inferences of causality using a cross-sectional research design. In addition, data collection was based on individual subjective questionnaire completion, which may have been biased. Finally, in our study, we surveyed 620 medical professionals in China, a small sample size that may affect certain statistical analyses or the ability to extend results to the general population. The next study will expand the sample size and scope of the survey and use a randomized controlled trial approach to assess the effects of social support, work environment, and risk perception interventions on medical staff resilience levels. We will also consider the impact of other factors on medical staff resilience levels and use longitudinal studies to validate the results.

Conclusion

The study confirms that risk perception, job stress and social support are significantly associated with the level of psychological resilience of medical staff during the COVID-19 pandemic. The positive effect of social support is the most prominent, while job stress and risk perception show negative effects. We further find that individual characteristic factors such as age, length of service, and job title moderate these influence pathways. These results suggest that in public health incidents, medical personnel with family harmony, coworker harmony, moderate risk perception, and work balance should be prioritized for emergency response work. A mental health protection system centered on resilience cultivation should be established to strengthen the social support system. This conclusion provides valuable insights for advancing theoretical research on factors influencing medical staff’s resilience.