Symptoms of anxiety, depression and fear in healthcare workers and non-healthcare workers undergoing outpatient COVID-19 testing in an urban Australian setting

Keywords: mental health, public health practice.

The psychological effects of the COVID-19 pandemic and the associated lockdown have been severe, with an increase in people reporting mental illness or not receiving ongoing treatment for previously existing mental illness (Pfefferbaum and North 2020; Rajkumar 2020). Generalised anxiety is more prevalent in areas with large COVID-19 caseloads and strict quarantine measures. A Spanish study reporting experiences early in the COVID-19 pandemic showed that 20% of the general community experienced significant symptoms of anxiety (Gonzalez-Sanguino et al. 2020), with higher levels (>30%) reported in similar Chinese studies (Wang et al. 2020a, 2020b). In Australia, in a sample of 1491 adults recruited through social media in April 2020, the prevalence of ‘any level’ anxiety, depression and stress was 21.2%, 38.3% and 27.8% respectively (Stanton et al. 2020). A similar online study of 5071 participants found rates that were nearly twice as high: anxiety, 49.8%; depression, 61.8%; and stress, 54.5% (Newby et al. 2020). In that study, fear of COVID-19 was specifically measured. Asked about the level of concern and worry about the possibility of contracting COVID-19, 66.4% were ‘a little’ or ‘moderately’ concerned and 25.7% were ‘very’ or ‘extremely’ concerned (Newby et al. 2020).

There is evidence that the burden of psychological morbidity is particularly high in health professionals, who may be faced with having limited resources to care for patients, a lack of protective equipment and concerns about infecting their own families with COVID-19 (Ahmed et al. 2020; Lai et al. 2020).

In the Australian health system, most mental health care is provided in a primary care setting (Royal Australian College of General Practitioners n.d.). Therefore, the task of caring for people with anxiety, depression and fear falls to primary health professionals, who are likely to care for healthcare workers (HCWs) as well as the general public. Understanding the extent and nature of psychological symptoms is therefore important to allow Australian primary care practitioners to prepare and plan for this role as the pandemic continues and to consider suitable interventions.

During the 4-month study period from June to September 2020, the incidence of COVID-19 was low in the state of New South Wales (NSW), Australia (population ~8 million). The highest daily total during this time was 23 cases on 10 August. This compares with the record highest daily total for the state of 213 cases on 27 March. At the time of the study, strict lockdown rules had been reduced and restrictions on gatherings were ongoing. Strict social distancing applied, and the wearing of masks was recommended but not mandated.

Westmead Hospital, in Western Sydney, NSW, is a referral centre for 1.5 million people and is the designated state infectious disease centre. Understanding the level of anxiety, depression and fear in people attending the clinic at Westmead Hospital will help determine the need for supportive care for clients and will provide insights into whether the needs of HCWs differ to those of non-HCWs attending for testing.

The aims of this study were to: (1) assess the levels of anxiety, depression and fear of COVID-19 in symptomatic people or close contacts (‘suspected COVID-19 cases’) attending for testing; and (2) identify associations between demographic groups and psychological morbidity; and (3) compare the prevalence of psychological symptoms between HCWs and laypersons attending the same testing clinic. A further aim was to use this information to help primary care practitioners prepare to care for people experiencing symptoms.

This cross-sectional study assessed psychological outcomes using validated self-administered web-based questionnaires in people attending for COVID-19 testing. The study was approved by the Western Sydney Human Research Ethics Committee (Reference: HREC 2020/ETH01071).

Participants were aged ≥18 years, triaged as meeting the criteria for testing at a COVID-19 testing clinic in a large public hospital and able to complete questionnaires in English on a mobile device. Only people with symptoms or close contacts of confirmed cases of COVID-19 (‘suspected cases’) were offered testing at the clinic at the time of the study. A participant information sheet was provided by nursing or administration staff on arrival. Interested participants scanned a QR code on the information sheet using their mobile phone to gain access to the online consent and study questionnaires.

Anxiety and depression were measured using the Hospital Anxiety and Depression Scale (HADS; Zigmond and Snaith 1983). Fear was measured using the Fear of COVID-19 Scale (Ahorsu et al. 2020; Soraci et al. 2020). The Fear of COVID-19 Scale has been validated in English and for online use in a New Zealand population (Winter et al. 2020), and has also been validated in an Australian multicultural population that included 42% HCWs (Rahman et al. 2020, 2021).

Descriptive statistics (mean ± s.d. or percentages) were used to describe participant characteristics and Chi-squared tests were used to test for differences between HCWs and non-HCWs. For anxiety and depression questionnaire data, logistic regression with a binomial distribution and log link function was performed to estimate relative risks; for fear of COVID-19, linear regression on the total scale score was performed using generalised linear models.

Further details about the methodology used in this study are provided in Supplementary Methods S1.

Between June and September 2020, 517 participants accessed the study questionnaires. Of these, 87 were excluded because <50% of items were completed, leaving responses from 430 (83.2%) participants in the final analysis.

Participant characteristics are presented in Table 1. The mean age of participants was 37.6 ± 12.6 years (range 18–83 years; median 35 years); 55.6% of the sample were female, 44.2% were male and 0.2% did not state gender. There were 151 HCWs (35.1% of the sample).

Table 1.  Participant demographics, with comparisons between health care workers (HCWs) and non-HCWs
Unless indicated otherwise, data are given as n (%). TAFE, technical and further education (non-degree diploma or certificate)

Mean anxiety scores, measured by the HADS, are presented in Table 2 for each group, with Table 3 summarising the results of univariable and multivariable analyses. Data were available for 430 participants. Overall, the mean anxiety score was 6.09 ± 4.41 and 279 of 430 participants (64.9%) had a ‘normal’ (‘non-case’ of anxiety) score, but 151 of 430 participants (35.1%) had a score ≥8, indicating a ‘case’ of anxiety according to the HADS definition, including 74 of 430 (17.2%) identified as having ‘borderline cases’ of anxiety and 77 of 430 (17.9%) identified as having ‘moderate/severe’ cases of anxiety. Gender, HCW status and having had a previous COVID-19 test were significant factors in univariable models and remained independently associated with the risk of anxiety in the multivariable model (Table 3). Male gender (relative risk (RR) 0.57; 95% confidence interval (CI) 0.43–0.75; P < 0.001) and being an HCW rather than a layperson (RR 0.53; 95% CI 0.39–0.72; P < 0.001) were associated with lower risk of anxiety, whereas having undergone a previous COVID-19 test was associated with a higher risk of anxiety (RR 1.30; 95% CI 1.02–1.66; P = 0.03). There were no significant two-way interactions.

Table 2.  Hospital Anxiety and Depression Scale (HADS) scores
Unless indicated otherwise, data are given as n (%). TAFE, technical and further education (non-degree diploma or certificate)

Table 3.  HADS anxiety and HADS depression analysis by demographic variables: univariable and multivariable analysis
HADS anxiety and depression scores were divided into two groups: non-cases versus borderline case/moderate-severe cases of depression or anxiety. Only variables with P ≤ 0.10 on univariable analysis were included in the multivariable model. CI, confidence interval; RR, relative risk; TAFE, technical and further education (non-degree diploma or certificate)

Mean depression scores, measured by the HADS, are presented in Table 2 for each group, with Table 3 summarising the results of univariable and multivariable analyses. Data were available for 430 participants. Overall, the mean depression score was 4.18 ± 3.60. Fewer ‘cases’ of depression (according to HADS criteria) were identified than ‘cases’ of anxiety; 348 of 430 participants (80.9%) had a ‘non-case’ of depression score and 82 (19.1%) had a ‘case of depression’ score, including 60 (14%) with a ‘borderline case’ of depression score and 22 (5.1%) with a ‘moderate/severe case’ of depression score. HCW status and level of education were significant factors in univariable models and remained independently associated with the risk of depression in the multivariable model (Table 3). Being an HCW rather than a layperson (RR 0.60; 95% CI 0.37–0.99; P = 0.04) was associated with a lower risk of depression. A lower level of educational attainment was also independently associated with the risk of depression (P = 0.01); compared with clients who had a postgraduate degree, there was a significantly higher risk of depression in those who had not completed high school (RR 3.84; 95% CI 2.11–7.03; P < 0.001) and those in the ‘Other/prefer not to say’ education category (RR = 2.77; 95% CI 1.35–5.68; P = 0.005). There were no significant two-way interactions.

Fear scores, measured by the Fear of COVID-19 Scale, are presented in Table 4. Data were available for 420 participants. The mean overall score for the cohort was 15.95 (maximum score 35, with higher scores indicating more fear). On analysis of the mean total score for fear of COVID-19, gender, HCW status and education were all statistically significant in univariable models. In the multivariable model, the mean fear of COVID score was significantly higher for females (17.5) than males (15.6; P < 0.001). There was a significant interaction between HCW status and level of education (P < 0.001). In HCWs, only the lowest level of education (high school or lower) was associated with a higher fear score (22.1) compared with a postgraduate degree (15.0; P < 0.001). In non-HCWs, all education levels below a postgraduate degree were associated with higher fear scores (P = 0.03 for undergraduate degree, technical and further education (TAFE) certificate or diploma; P = 0.02 for high school education or lower).

Table 4.  Fear of COVID-19 (total score) analysis by demographic variables: univariable and multivariable analysis
Only variables with P ≤ 0.10 on univariable analysis were included in the multivariable model. CI, confidence interval; TAFE, technical and further education (non-degree diploma or certificate)

This study evaluated anxiety, depression and fear in a cohort of participants attending a hospital clinic for ambulatory COVID-19 testing. Participants were symptomatic or were close contacts of a confirmed case of COVID-19. Because the clinic was based in a tertiary hospital, 35% of attendees were HCWs. Although the number of HCWs was small (n = 151), the data provide some insight into the mental health of HCWs, as well as that of the general community. During the study period, the number of new cases of COVID-19 in the community remained low. Despite low levels of COVID-19 in the community, the prevalence of symptoms of anxiety and fear of COVID-19 was high.

In this study, HCWs had fewer anxiety symptoms than non-HCWs. HCWs had similar mean scores for fear of COVID-19 to non-HCWs (adjusted mean ‘cases’ of 16.6 vs 16.4 respectively; not significant on multivariable analysis). Previous research demonstrated high levels of fear and anxiety in HCWs during the COVID-19 pandemic in China, regardless of whether they were deployed to frontline positions (Liu et al. 2020; Wu et al. 2020). It is likely that HCWs in the present study were less concerned because the number of cases of COVID-19 was low and the health system was not overwhelmed, as it was at the time of previous research in China. It is also possible that the higher level of health literacy and greater familiarity with the clinical setting of HCWs relative to non-HCWs may lead to less anxiety among HCWs attending the clinic, although the HADS questionnaire asks participants to report symptoms over the preceding week, rather than just present symptoms. Although HCWs had similar levels of fear about COVID-19 as non-HCWs overall, their level of anxiety was low (and was lower than non-HCWs and the population normative levels.) This provides some evidence that HCWs appeared to be maintaining aspects of their mental health relatively well at this more ‘chronic’ stage of the pandemic. It is also possible that HCWs are well supported with current interventions or that the experience of HCWs is different in the Australian health setting than in other parts of the world.

The assessment of anxiety in this study demonstrated more anxiety symptoms in women than men (mean scores 6.8 vs 4.1 respectively; P < 0.001), and this is consistent with previous studies. The mean HADS score in a baseline cohort study in the UK was 6 for women and 5 for men (Breeman et al. 2015). In addition, the prevalence of anxiety symptoms measured by the HADS was higher than expected in women, with the present study showing 23% of women had moderate/severe anxiety symptoms measured by HADS, compared with 19% in the baseline cohort study of Breeman et al. (2015). For men, there was a lower than expected prevalence of anxiety symptoms measured by HADS the present study, with 11.6% of men showing moderate/severe anxiety symptoms compared with 12.5% in the normative study (Breeman et al. 2015). Research in the general population early in the pandemic found a prevalence of anxiety symptoms of 21% in women in Spain (Gonzalez-Sanguino et al. 2020), measured by the Generalized Anxiety Disorder 2-item questionnaire (GAD-2). Prevalence in China was reported to be between 6.3% (Wang et al. 2021) and 36.4% (Wang et al. 2020a) in general populations during the first wave of the pandemic. It must also be acknowledged that all participants in the present study had symptoms that could represent COVID-19 or were close contacts of a confirmed case, and this, along with their presence in the testing clinic, could have contributed to the measured anxiety symptoms.

This study found a relatively low prevalence of moderate/severe symptoms of depression measured by HADS. Although the mean depression score of 4.2 was higher than the reported normative score of 3 (Breeman et al. 2015), the prevalence of moderate/severe symptoms of depression of 5.3% in men and 5% in women (non-significant difference) was lower than the normative prevalence of 6.9% for both genders (Breeman et al. 2015).

Because it is a new measure, few studies have previously used the Fear of COVID-19 Scale. In the present study, the mean fear of COVID-19 score was 17.5 in women and 15.6 in men (P < 0.001). This is similar to the score of 16.9 in a predominantly female Italian population at the height of the first phase of the pandemic (Soraci et al. 2020). This indicates that the population in the present study has a high level of fear (equivalent to the early days of the pandemic in Europe), despite the low number of COVID-19 cases during the study period. Similar mean scores for the individual items and similarly higher levels of fear in women were found in an Israeli study (Tzur Bitan et al. 2020). In a New Zealand population, mean fear of COVID-19 scores of 18.3 and 15.6 were found in two mixed male–female groups (Winter et al. 2020). The higher score was found during the strictest lockdown phase and the lower as restrictions began to ease (Winter et al. 2020). In an Australian population, a higher level of COVID-19-related fear was noted in women and younger people, similar to the findings in the present study (Rahman et al. 2020).

Level of educational attainment was strongly associated with anxiety symptoms and fear of COVID-19. This demonstrates the importance of specific and targeted strategies to support this vulnerable group that may have associated low health literacy.

The strengths of the study are the relatively large sample size and the use of standardised and validated measurement tools. The main limitation of this study is that it was conducted in a COVID-19 testing clinic with symptomatic participants or close contacts of COVID-19 cases, so it is unclear whether the results can be applied to the general population. However, it provides insight into the mental health of symptomatic people in the setting of COVID-19, and it provides some information about the well-being of people working in the healthcare setting. Because the number of HCWs in the study was relatively small, the generalisability of these results cannot be assumed. Also, detailed analysis of the HCW by role was not undertaken due to the small numbers in the subgroups. Information about comorbid or an underlying diagnosis of mental illness was not collected, so the relative contribution of the pandemic to the measured symptoms of anxiety and depression is unclear. The study response rate was challenging to reliably quantify due to infection control measures necessitating the use of a web-based platform linked from QR codes in the clinic. Although the incompletion rate of forms (16.8%) was quite low, this may have led to a selection bias in the reported results. It may be anticipated that the level of anxiety would be high in people with ‘suspected’ COVID-19. In addition, the incidence of positive COVID-19 cases fluctuated during the study period and was low overall, so the results may not be applicable to high-caseload areas.

NSW (especially Western Sydney, the study setting) is experiencing a further wave of COVID-19 cases during the second half of 2021 in the setting of low vaccination rates. Daily case numbers have exceeded those during the first wave and lockdown restrictions have returned. Further research is needed to explore how fear of COVID-19 is affecting people during this quite different phase of the pandemic. The data presented in this study provide a point of comparison for current and future research.

The pandemic has caused a significant psychological impact worldwide. This study provides an indication that many Australians were also significantly affected, even in a community where infection rates were low. This stresses the importance of ongoing interventions to support psychological well-being in both HCWs and the wider community. In Australia, most of this burden is carried by primary care practitioners; however, due to the prevalence of symptoms, additional support with large-scale public health interventions may also be required. Strategies to address anxiety could be targeted to women and people who had had lower levels of formal education, as well as to the community in general. The point of testing provides a contact opportunity and may be feasible for interventions now that the testing process has become streamlined. HCWs were found to have relatively lower levels of anxiety and fear than the general community attending the same clinic. Further research is needed to understand whether this is related to the relatively low COVID-19 burden in Australia during the study period, whether the current supports for HCWs are effective or whether HCWs are responding differently in the Australian healthcare system compared with other parts of the world.

This study has demonstrated high levels of anxiety and fear of COVID-19 in laypersons attending ambulatory COVID-19 testing, despite a very low likelihood of a subsequent positive result. There is a strong correlation between fear and anxiety with female gender, people with a lower level of education and individuals attending for repeat testing. The overall level of fear in various subgroups was similar to the alarmingly high results found in Europe in the early stage of the COVID-19 pandemic. HCWs had a lower level of anxiety and fear than members of the general community, and depression scores were comparable with the wider population across both groups. This provides some reassurance that, in an Australian setting with low rates of COVID-19, health professionals are experiencing relatively low levels of anxiety and depression symptoms. Further research could focus on interventions to address the issues identified in the general community, which may include innovative ways to introduce support for mental health at the time clients attend for testing and support for management in primary care.

The data supporting this study are available in the Mendeley Data Repository (doi:10.17632/5psjcfssc7.1).

The authors declare no conflicts of interest.

M. L. Marinovich is supported by funding from a National Breast Cancer Foundation Investigator Initiated Research Scheme grant (IIRS-20-011).