COVID-19-related stressors are likely to contribute to nurses' psychological well-being. For example, a prior study found that perceived risk of contracting COVID-19 was associated with anxiety symptoms, frontline nurses in China during COVID-19 pandemic (Cui et al., 2021). Working at hospitals and clinics during COVID-19 pandemic might increase nurses' worries about passing infection to family members being due to their jobs, which in turn affect healthcare workers' well-being (Li et al., 2020; Walton et al., 2020). It has also been found that healthcare workers' worries about insufficient personal protective equipment were associated with higher distress among healthcare workers in the United States (Hennein et al., 2021) and Korea (Han et al., 2021). Relevant studies are limited in the Hong Kong context. Therefore, we speculated that COVID-19-related worries would also be associated with more anxiety symptoms among local nurses in Hong Kong.

When facing highly stressful events like the COVID-19 pandemic, having different aspects of coping resources available could contribute to nurses' well-being (Labrague and De Los Santos, 2020). A recent review has supported that personal coping resources and work-related resources/risk factors are important determinants of psychological well-being among employees during the COVID-19 crisis (Rigotti et al., 2021), including healthcare workers. In this study, we aimed to examine the roles of interpersonal and work-related coping resources in anxiety symptoms among nurses in Hong Kong.

Regarding interpersonal resources, emerging research has started to examine the role of perceived stigma in healthcare workers' well-being during the COVID-19 pandemic. Due to the job nature, it is inevitable for healthcare workers to have the close contact with suspected and diagnosed cases of COVID-19 in the pandemic situation. Healthcare workers are commonly regarded as individuals with high risks for infecting and transmitting the virus. In the contexts of prior pandemic situations like the Severe Acute Respiratory Syndrome (SARS) and the Middle East Respiratory Syndrome (MERS), perceived stigma has been found to be a significant contributor of poor psychological well-being among healthcare workers (Gupta and Sahoo, 2020). Specific to the context of COVID-19, perceived stigma was associated with adverse mental health outcomes (e.g., psychological distress, probable depression) among healthcare workers in Bangladesh (Khan et al., 2021) and the US (Hennein et al., 2021). Health professionals' exposures to stigma or discriminatory behaviors may lead to gradual physical and psychological deterioration (Ramaci et al., 2020). In Hong Kong, healthcare workers are generally highly respected professions (Schoeb, 2016). It is important to examine how perceived stigma of being a healthcare worker contribute to local nurses' well-being in the context of COVID-19.

Research also suggests that work-related coping resources (e.g., level of job satisfaction, pandemic control measures at the institutions) can be associated with well-being among healthcare workers. Studies have found that individuals who are satisfied with work are more likely to find meaning in nature of work, which facilitate well-being during the COVID-19 pandemic (Hamama-Raz et al., 2021). Another study found that higher work satisfaction was associated with perceptions of positive changes (e.g., secondary posttraumatic growth) from the work experience during the pandemic among paramedics and nurses in Poland (Ogińska-Bulik et al., 2021a). The protective role of work satisfaction in anxiety symptoms among nurses should be further examined. On the other hand, workplace arrangements matter in employees' well-being during the pandemic. For example, implementation of COVID-19-related accommodating measures (e.g., introduction of reliable information resources, preventive measures to reduce risk of infection at workplace) were associated with lower psychological distress and better job performance among a heterogeneous sample of Japanese employees from multiple industries (Sasaki et al., 2020). Having transparent and timely policies for prevention of nosocomial infections (e.g., healthcare-associated infections) at the hospitals/clinics was associated with lower risks for probable depression and anxiety among healthcare workers in the US (Hennein et al., 2021). Whether satisfaction with workplace pandemic control guidelines could contribute to anxiety symptoms among local nurses has yet to be explored. Based on the above-mentioned literature, we expected that work satisfaction and satisfaction with workplace pandemic control guidelines would be associated with lower anxiety symptoms among Hong Kong nurses during the COVID-19 pandemic.

This study examined the psychosocial correlates of anxiety symptoms among nurses in Hong Kong amid the COVID-19 pandemic. Based on the aforementioned studies, we hypothesized that higher levels of COVID-19 worries and perceived stigma, plus lower levels of work-related resources (work satisfaction, satisfaction with workplace pandemic control guidelines) were associated with higher anxiety symptoms.

The nurses working in either public or private service provision in different settings (including inpatient, outpatient, outreach service in community setting) were eligible for this study. Those who were nursing trainees and retired nurses were excluded from the study sample. All registered members of the Association of Hong Kong Nursing Staff (n = 16,500), the labor union of nurses in Hong Kong, were approached and invited to this study using their email contacts. The self-administered questionnaire was distributed to the nurses in an internet-based link along with an invitation email. An information sheet about the study was included at the beginning of the questionnaire, followed by an electronic consent form. The participants who agreed to join the study filled in the questionnaire on their own electronic devices. A reminder for participation into the survey was sent 2 weeks after the first invitation email. Responses from 1,566 participants were collected via the online platform. Among them, 56 did not fulfill the eligibility criteria (e.g., not being a nurse, being retired, or being a nursing student). Therefore, only 1,510 valid responses representing working nurses included both worked as full-time or part-time in health care settings were retained in the analyses.

Comparison of sample characteristics was made with statistics of nurse population in Hong Kong to explore the potential selection bias. Although the response rate was low (1,510/16,500), characteristics of this sample were matched to nurse population in Hong Kong according to the latest statistics from Department of Health (Department of Health, 2016).

Upon completing the questionnaire, participants received HKD$50 (approximately USD$6.43 for compensation of their time. The study was conducted between August 8, 2020—September 22, 2020 (when there were 5,059 patients diagnosed and 105 dead from COVID-19 in Hong Kong). The study protocol was approved by the Research Ethics Committee (CUHK-NTEC CREC) at the first author's institution (Protocol no. CRE-2020.073).

Descriptive statistics and Pearson correlations among the major variables were computed. Internal consistencies of the scales were indicated by their corresponding Cronbach's alphas. Hierarchical regressions were conducted to examine the associations between the independent variables and anxiety symptoms. The sequence of entering independent variables followed the suggestions from prior studies on individuals' well-being in response to emerging infectious diseases and other traumatic events (Yeung et al., 2016, 2021). In the first block, background variables showing significant associations with anxiety symptoms in bivariate correlations were entered. In the second and third blocks, COVID-19 worries, and perceived stigma were entered in the model, respectively. In the last block, work-related coping resources (i.e., work satisfaction, satisfaction with workplace pandemic control guidelines) were entered. The analyses were conducted using SPSS 26.0.

Expecting a small-to-medium effect size (f ² = 0.05) in the association between the independent variables and anxiety symptoms in the hierarchical regression analysis, we needed at least 515 participants to achieve a statistical power of 0.95 at p = 0.05 (G^*Power 3.1.2). With our sample size (N = 1,510), we should be able to detect the expected effect size with sufficient statistical power.

Most of the participants were aged between 30 and 39 (36.8%) and married (50.7%). On average, they have been working in the profession for around 8.9 years (SD = 8.60). About one-fifth of the participants were specifically taking care for diagnosed and suspected cases of patients with COVID-19 (‘dirty team”) (18.5%) and reporting at least a moderate level of anxiety (GAD ≥ 10) (17.2%) (Table 1).

The correlation analysis results showed that higher levels of COVID-19-specific worries (infecting with COVID-19, family members infecting with COVID-19 due to the participants' nursing duties, insufficiency of protective equipment at workplace), higher perceived stigma, lower work satisfaction, and lower satisfaction with workplace pandemic control guidelines were correlated with higher anxiety symptoms (rs ranged from 0.12 to 0.39, p < 0.05) (Table 2). Regarding socio-demographic variables, younger age, fewer years in the profession, not being married, having children (rs ranged from −0.15 to −0.09, ps < 0.001), and working in the “dirty team” were correlated with more anxiety symptoms (r = 0.12, p < 0.001). Those variables were included as covariates in the subsequent regression analysis. Other socio-demographic variables reported non-significant correlations with anxiety symptoms (ps > 0.05, not tabulated).

Given that the independent variables were moderately correlated with each other, the independent variables were checked for multicollinearity in the regression analysis. We did not find any variables reporting a variance inflation factor (VIF) ≥ 4, supporting the absence of multicollinearity. In Block 1, the background variables explained 3.3% of variance in anxiety symptoms. Specifically, younger age (β = −0.12, p < 0.01) and working in the “dirty team” (β = 0.10, p < 0.01) were significantly associated with higher anxiety symptoms. In Block 2, higher levels of worries about contracting COVID-19 from work, about family members contracting COVID-19 due to their nursing duties, and about insufficient protective equipment at workplace (βs ranged from 0.13 to 0.25, ps < 0.001) were significantly associated with higher anxiety symptoms. In Block 3, higher perceived stigma of being a healthcare worker was associated with higher anxiety symptoms (β = 0.20, p < 0.001), explaining an additional 3.7% of variance. In Block 4, work-related coping resources variables explained an additional 3.8% of the variance in anxiety symptoms. Only work satisfaction was associated with lower anxiety symptoms (β = −0.21, p < 0.001). With all the independent variables, the overall model explained 30.4% of variances in anxiety symptoms (Table 3).

This study had several limitations. First, it was a cross-sectional study so that the tested relationships were not causal. Future studies could examine how the changes in the independent variables could predict anxiety symptoms temporally using longitudinal designs. Second, all of the measures were self-reported, which might be subject to recall bias. However, the use of self-reported measures is common among many studies conducted during the COVID-19 pandemic (Salari et al., 2020) due to its convenience, low cost, and ease of administration at pandemic situations. Third, we recruited nurses through nursing associations in Hong Kong, where self-selected bias might be apparent. Comparable to response rates in some prior studies targeting at healthcare workers conducted online during the COVID-19 pandemic (Alenazi et al., 2020; Ammar et al., 2020), our response rate was low at 9.2% (1,510/16,500). Despite the low response rate, characteristics of this sample were matched to nurse population in Hong Kong according to the latest statistics from Department of Health (Department of Health, 2016). Readers should be noted that the findings might not be fully generalizable to the nurses and healthcare professionals in other countries with different medical systems. Fourth, the model only explained a moderate proportion of variance in anxiety symptoms. Other factors may be at play. For example, personality attributes (e.g., resilience, Type D personality) (Labrague and De Los Santos, 2020; Tuman, 2021), coping strategies (Si et al., 2020; Özçevik Subaşi et al., 2021), and impact of the pandemic on family roles have been found to be important contributors to people's well-being during COVID-19 (Cipolotti et al., 2020; Luceño-Moreno et al., 2020). Considering these variables could allow a more comprehensive capture the contributors to people's mental health outcomes in response to the pandemic. Fifth, this study targeted on an underserved population in the literature. Some scales and items (e.g., COVID-19 distress, work-related stress) were newly developed for the context of COVID-19, which were yet to be fully validated in this population. Specifically-developed items were also commonly used as predictors of mental health outcomes among different Asian populations during the COVID-19 pandemic (Choi et al., 2020; Yeung et al., 2020). Even our self-developed scales reported satisfactory psychometric properties, validating our findings by measuring the same concepts (e.g., Work Satisfaction Scale, perceived stigma for healthcare workers) (Traynor and Wade, 1993; Mostafa et al., 2021) or similar concepts (e.g., COVID-19 organizational support) (Zhang et al., 2020) with other standardized scales is worthwhile.

This study highlighted that COVID-19 worries, perceived stigma, and work satisfaction might be prominent contributors to anxiety symptoms among those nurses. The World Health Organization recommends that healthcare workers should maintain self-care, be attentive to their own mental health problems, and reflect on how their experiences may influence themselves and their loved ones. Identifying and instituting effective treatment strategies to improve psychological outcomes for nurses is essential. Practically, our findings also implied that alleviating COVID-19 worries, addressing stigma toward healthcare workers, and increasing work satisfaction might be important intervention strategies to reducing anxiety symptoms among the nurses amid the COVID-19 pandemic. Development of digital interventions targeting at healthcare workers' well-being has also been emerging. Recently in the United Kingdom, a free digital education package has been developed to provide valuable information to enhance healthcare workers' skills in coping with social stigma, self-care, and social support seeking during the pandemic (Blake et al., 2020). On the other hand, experimental studies also indicated that expressive writing interventions could be effective in improving psychological health among Italian healthcare workers during the COVID-19 pandemic (Procaccia et al., 2021). Tailoring an intervention for local healthcare workers based on our findings, it is worth exploring if similar mobile applications providing additional modules to allow individuals to reflect on positive work experience and meaning derived from job duties, plus express negative emotions related to stigmatization experience would facilitate well-being.