Beyond the physical risk: Psychosocial impact and coping in healthcare professionals during the COVID‐19 pandemic

Abstract Aims and objectives This study aimed to examine the psychosocial impact and identify risk factors for poor psychosocial outcomes in healthcare professionals during the Coronavirus disease 2019 (COVID‐19) pandemic in Cyprus. Background Healthcare professionals are in the forefront of the COVID‐19 pandemic facing an unprecedented global health crisis, which can have consequences on their psychosocial health. There is a need to identify risk factors for poor psychosocial outcomes to inform the design of tailored psychological interventions. Design Cross‐sectional online study. Methods A total of 1071 healthcare professionals completed self‐report questionnaires. Measures included sociodemographic information, COVID‐19‐related characteristics, quality of life (Brief World Health Organization Quality of Life; WHOQOL‐Bref), anxiety (Generalized Anxiety Disorder‐7; GAD‐7), depression (Patient Health Questionnaire‐8; PHQ‐8), occupational burnout (Copenhagen Burnout Inventory; CBI), and coping (Brief Coping Orientation to Problems Experienced; Brief COPE). This article follows the STROBE reporting guidelines. Results The prevalence of moderate to severe anxiety and clinically significant depression was 27.6% and 26.8%, respectively. Significant risk factors for poor psychological outcomes included being female, being a nurse or doctor (vs non‐medical professional), working in frontline units (inpatient, intensive care), perceptions of inadequate workplace preparation to deal with the pandemic, and using avoidance coping. Depression and occupational burnout were significant risk factors for poor quality of life. Conclusion The findings suggest several individual, psychosocial, and organisational risk factors for the adverse psychological outcomes observed in healthcare professionals during the COVID‐19 pandemic. Relevance to clinical practice This study highlights the urgent need for screening for anxiety and depression and psychological interventions to combat an imminent mental health crisis in healthcare professionals during the COVID‐19 pandemic. Pandemic response protocols and public health initiatives aiming to improve and prevent mental health problems in healthcare professionals during the current and future health crises, need to account for the various factors at play.


| INTRODUC TI ON
to 2019, approximately 17% of Cypriots had to pay to access the public health system and/or services in the private health sector (Theodorou et al., 2012). The COVID-19 pandemic further complicated this transitional period, causing delays in the introduction of certain healthcare services and exacerbating the shortage of healthcare staff in public hospitals (WHO Regional Office for Europe et al., 2020). Furthermore, organisational changes were made to cope with the pandemic emergency situation, whereby an entire hospital (out of six public hospitals) was converted into a COVID-19 treatment centre suspending most other services, many clinics within hospitals were converted to COVID-19 wards, healthcare professionals from various departments were assigned to the treatment of COVID-19 patients, and non-essential outpatient appointments and non-emergency surgeries were postponed (Press & Information Office, 2020a). At times, the increasing infection rates among inpatients and hospital staff has led to temporary suspension of services of major public hospitals and clinics. A significant proportion (21.4%) of the 873 confirmed cases in Cyprus between March and May 2020 were healthcare professionals, of which approximately 50% were nurses (Quattrocchi et al., 2020).
The healthcare challenges posed by the COVID-19 pandemic coupled with the organisational transition placed increased pressure on healthcare professionals; they have been in the forefront of the COVID-19 pandemic response and many of them have not experienced a global health crisis of these proportions before. Additional stressors are introduced in an already hectic setting, including the increased infection risk, global lack of protective and medical equipment during the early stages, lack of a very effective COVID-19 treatment (with a few treatment modalities approved) (Beigel et al., 2020;RECOVERY Collaborative Group et al., 2021;Sterne et al., 2020), extended working hours, caring for severely ill patients, and taking life and death decisions among others; all while being isolated from family and friends (Santarone et al., 2020). These working conditions have put healthcare professionals at greater risk for psychological distress, fatigue, burnout, and stigma (Arora et al., 2020;Muller et al., 2020;Paiano et al., 2020;Pappa et al., 2020). The WHO has therefore urged healthcare professionals to care for their mental health and use positive coping strategies to deal with the increased demands during this pandemic (WHO, 2020). Research on the most suitable ways to help our healthcare workforce cope with the psychological impact of the COVID-19 pandemic has been identified as an immediate priority (Zaka et al., 2020).

| BACKG ROU N D
During the SARS pandemic, healthcare workers experienced posttraumatic stress, depression, and anxiety symptoms especially if mental health crisis in healthcare professionals during the COVID-19 pandemic.
Pandemic response protocols and public health initiatives aiming to improve and prevent mental health problems in healthcare professionals during the current and future health crises, need to account for the various factors at play.

K E Y W O R D S
COVID-19, healthcare professionals, quality of life, anxiety, depression, occupational burnout, coping, SARS-CoV-2, Pandemic, mental health What does this paper contribute to the wider global clinical community?
• There are notable levels of anxiety and clinically significant depression in healthcare professionals during the COVID-19 pandemic.
• Risk factors for poor QoL, anxiety, depression, and occupational burnout include female gender, being a nurse or doctor (vs non-medical professional), working in frontline units (inpatient wards, ICUs), inadequate workplace preparation to deal with the pandemic, and engaging in avoidance coping, while the presence of depression and occupational burnout further impedes QoL in healthcare professionals.
• The findings can help identify healthcare professionals in need for psychological support and in the design of tailored psychological interventions, public health initiatives, and future protocols to help healthcare professionals cope with the psychosocial burden of COVID-19 and build resilience against future health crises. they worked in high-risk clinical settings (Wu et al., 2009). There is now a growing body of research focusing on the psychological consequences of COVID-19 in healthcare professionals, with the majority originating in China. In addition to the physical risks, these studies have found increased levels of depression, anxiety, insomnia, and distress in healthcare professionals during the COVID-19 pandemic (Lai et al., 2020;Xiao et al., 2020;J. Zhang et al., 2020). Several factors have been reported to increase the risk for psychological symptoms in healthcare workers, including being a nurse, female gender, and frontline work (Muller et al., 2020;Pappa et al., 2020). The increased workload and psychological distress during the COVID-19 pandemic can further decrease work engagement in frontline nurses potentially influencing the provision of quality of care (Zhang et al., 2021). To our knowledge, there is limited evidence on the psychosocial impact of COVID-19 on healthcare professionals in Cyprus, a country that has no extensive experience in dealing with pandemics of this scale and lacks psychological support services tailored to the needs of healthcare professionals.
It is also important to identify the underlying psychological processes, which can serve as protective factors against the negative psychosocial consequences of the pandemic in healthcare professionals. One such factor is coping, which is motivated action underlined by self-regulatory processes that helps people adapt to stressful situations (Carver, 1997). The SARS literature has highlighted the role of maladaptive coping skills, including substance use, self-blame, and avoidance as risk factors for psychological symptoms in healthcare professionals even long after the outbreak was resolved (Maunder et al., 2008). Healthcare professionals working in different roles and settings during the SARS epidemic used diverse maladaptive coping strategies. For example, a study in Hong Kong reported that higher levels of distress in emergency doctors dealing with SARS were associated with greater use of venting as a coping strategy, whereas higher levels of distress in nurses were associated with greater use of behavioural disengagement and self-distraction (Wong et al., 2005). Coping and how it relates to the psychological responses and QoL of healthcare professionals during COVID-19 have received little attention. A study conducted in Italy has found that a positive attitude held by healthcare professionals during COVID-19 was associated with lower perceived stress, whereas seeking support and avoidance coping were associated with higher perceived stress (Babore et al., 2020). A qualitative study, on the other hand, has highlighted the benefit of having a support system in terms of stress in healthcare workers during the COVID-19 pandemic (Eftekhar Ardebili et al., 2020).
There is need for more evidence on how the use of certain maladaptive coping strategies may put healthcare professionals at risk for mental health problems and poor QoL that will help inform a strategy to enhance their coping skills and support them during this pandemic. Such research will satisfy the urgent need, as highlighted by a recent call for action paper, to provide psychological interventions for healthcare professionals tailored to their current needs (Zaka et al., 2020). The focus of much of the research has been on anxiety and depression as psychological outcomes. Quality of Life

| Setting
The study consisted of an anonymous online survey administered between May 25th and October 27th, 2020 in Cyprus.

| Participants
The study followed a convenience sampling approach and included healthcare professionals of the public and private sector defined as physicians, nurses, allied health professionals, and health scientists and academics involved in the Cypriot government's efforts to combat COVID-19. Exclusion criteria included: (a) healthcare professionals who were not working or were absent (e.g., abroad, medical leave, sabbatical leave) during the whole duration of the pandemic, (b) inability to consent without the help of a third party, (c) inability to understand Greek. Participants were recruited through mass e-mails from professional organisations, including the Cyprus Medical Association (CYMA) and the Cyprus Nurses and Midwives Association (CYNMA), social media platforms, and national newspapers. Prospective participants were provided with a study link containing information about the study, the electronic consent form, and the survey. The survey was administered through the Research Electronic Data Capture (REDCap) (Harris et al., 2009)

| Measures
The survey consisted of three sections of self-report questionnaires.
The first section included sociodemographic and COVID-19-related characteristics, including age, gender, family status, education, profession, years of experience in clinical settings, experience in crisis situations in clinical settings, work setting and department during the pandemic, contact with COVID-19 patients (frontline vs nonfrontline staff), adequacy of COVID-19 preparation in their workplace, need for self-isolation during the pandemic, and whether they were diagnosed with COVID-19. The second section included questionnaires relating to psychosocial functioning and QoL. QoL was as-  (Spitzer et al., 2006). Participants are asked to rate how often they were bothered by certain issues during the past two weeks using a three-point Likert scale of seven items. The total score can range from zero to 21 with higher scores indicating higher anxiety levels. Scores of 5, 10, and 15 may represent cut-off points for mild, moderate, and severe anxiety, respectively (Spitzer et al., 2006).
Participants are asked to rate how often they were bothered by certain issues during the past two weeks using a three-point Likert scale of eight items. The total score can range from zero to 24 with higher scores indicating higher depression levels. A total score of zero to 4 represents no significant depressive symptoms, 5 to 9 represents mild symptoms, 10 to 14 represents moderate symptoms, 15 to 19 represents moderately severe symptoms, and 20 to 24 represents severe symptoms. A cut-off point of larger or equal to ten defines current depression (Kroenke et al., 2001(Kroenke et al., , 2009. Occupational burnout was assessed using the relevant scale of the Greek translation of the Copenhagen Burnout Inventory (CBI) (Kristensen et al., 2005;Papaefstathiou et al., 2019). Participants are asked to rate the degree to which they experience exhaustion due to their work using a five-point Likert scale of seven items. The mean score can range between 0 and 100, with higher scores indicating greater occupational burnout. The third section included the Greek translation of the Brief Coping Orientation to Problems Experienced inventory (Brief COPE) (Carver, 1997) which assesses individuals' coping strategies as self-regulatory processes. Participants are asked to indicate the degree to which they used each strategy when they experienced stress during the past two weeks using a four-point Likert scale. The instrument measures 14 coping reactions (two items each), including active coping, planning, positive reframing, acceptance, humour, religion, using emotional support, using instrumental support, selfdistraction, denial, venting, substance use, behavioural disengagement, and self-blame. Mean scores are obtained for each subscale which can range between 0 and 3, where higher scores indicate greater use of the specific coping strategy. Exploratory factor analysis may be performed on the subscales of the Brief COPE to explore the underlying second-order factors (Carver, 1997).

| Statistical analyses
To estimate a statistically sufficient sample size for the study, a priori power analysis using G-Power (Faul et al., 2007) was conducted. The sample size was calculated with the aim of detecting a medium effect size (Cohen, 1988) in the main outcomes (quality of life, anxiety, depression, burnout) and at α = 0.05. Approximately 179 participants would be needed to achieve a power of 90% when using linear multiple regression analysis with a maximum of 17 predictor variables. These calculations were used as a reference point to indicate the minimum number of participants required for a statistically sufficient sample size and there was no maximum threshold.
Data from REDCap were exported to IBM SPSS 26, where all analyses were performed. Details of the missing value analysis are provided in Supplementary File 2. Complete case analysis was selected since missingness was associated with independent variables, producing no great risk for bias (Hughes et al., 2019). Exploratory factor analysis (EFA) was conducted on the BRIEF COPE subscales using Principal Axis Factoring extraction with orthogonal (varimax) rotation to identify second-order factors. The appropriateness of EFA was confirmed by examining the factorability, the Kaiser-Meyer-Olkin measure of sampling adequacy, Bartlett's test of sphericity, and the values of the diagonals of the anti-image correlation matrix. Factors were extracted based on the Kaiser rule (eigenvalues > 1) and the scree plot. A cut-off of >0.4 was selected as the minimum factor loading (Field, 2009). Composite scale scores were calculated based on the mean score of the subscales that loaded onto a given factor. Descriptive analyses were performed for sociodemographic, COVID-19-related variables, anxiety, and depression.
Multiple linear regressions were performed to examine the demographic, COVID-19-related, and psychosocial predictors of anxiety, depression, occupational burnout, and QoL outcomes. In the regression models for QoL outcomes, anxiety, depression, and burnout were also entered as predictors. Predictors were entered in the models using the entry method and if they were significantly associated with the outcomes in bivariate correlations. Dummy variables were used where appropriate. To examine differences in the use of coping skills among professionals working in different contexts and settings, we conducted independent samples t-tests and one-way ANOVA with post hoc tests (Tukey HSD for equal or Games-Howell for unequal variances) based on 95% confidence intervals. The Welch test was performed when the homogeneity of variances assumption was violated. A p-value < .05 (two-tailed) was considered to indicate statistical significance.

| Ethical statement
The study was conducted according to the guidelines of the Declaration of Helsinki and has received ethical approval from the Cyprus National Bioethics Committee (approval number: 2020.01.106). Participants completed an online consent form and all data collected were anonymous and confidential.

| RE SULTS
Results of missing value analysis and differences between complete and incomplete cases are available in Supplementary File 2.
The total number of participants was 1071 and 73% were female,

| Exploratory factor analysis of the Brief COPE
EFA of the Brief COPE yielded a three-factor solution based on eigenvalues >1, which explained 41.65% of the variance. The scree plot (see Supplementary File 3) confirmed the three-factor structure.
The subscales of humour, religion, and substance use did not load on any of the factors at >0.40 and had low communalities (<0.20), thus were removed from the model. This resulted in a better three-factor solution, which explained 48.92% of the variance. Table 2 contains the results of EFA with the rotated factor loadings. The three extracted factors represented 'Approach' (active efforts to deal with the problem), 'Support-seeking' (seeking support from the environment), and 'Avoidance' (avoiding dealing with the problem) coping.

| Predictors of anxiety, depression, and occupational burnout
The correlations between variables that informed the multiple linear regressions are available in Supplementary File 4. The coefficients of multiple regressions for anxiety, depression, and occupational burnout are presented in Table 3. Considering the sociodemographic predictors, being female was associated with higher levels of anxiety and occupational burnout (beta = 0.07, p < .05), being a medical doctor (vs non-medical professional) was associated with greater occupational burnout (beta = 0.08, p < .05), while being a nurse or midwife (vs non-medical professional) was associated with greater depression and occupational burnout (betas = 0.08 and 0.10, p < .05).

| Predictors of quality of life
The coefficients of multiple regressions for QoL outcomes are presented in Table 4. Being female was associated with poorer psychological health (beta = −0.07, p < .05), older age was associated with poorer overall health (beta = −0.26, p < .001), and being single and being a medical doctor (vs non-medical professional) were associated with poorer social relationships (beta = 0.14, p < .001 and beta = −0.09, p < .05, respectively). Two COVID-19-related predictors were significantly associated with QoL outcomes. Inadequate workplace preparation to deal with the pandemic was associated with poorer overall QoL (beta = 0.07, p < .05), social relationships (beta = 0.07, p < .05), and environmental QoL (beta = 0.10, p < .001), while being diagnosed with COVID-19 was associated with poorer environmental QoL (beta = −0.06, p < .05). The single most common psychosocial predictor of poor QoL across all outcomes was depression (betas = −0.29 to −0.52, p < .001). Occupational burnout was associated with poor QoL in all outcomes except psychological health (betas = −0.11 to −0.19, p < .001). Anxiety was only associated with poor environmental QoL (beta = −0.10, p < .05). Using avoidance coping in greater extent was associated with poorer psychological health (beta = −0.14, p < .001) and social relationships (beta = −0.07, p < .05).
A summary of the statistically significant predictors across all psychosocial outcomes is provided in Supplementary File 5.

| Differences in the use of coping skills
The results of comparisons in the use of coping skills based on sociodemographic and COVID-19-related characteristics are available in Table 5. Women used approach and support-seeking coping to a greater extent than men (mean difference = −0.11, p = .002, 95%CI TA B L E 1 Sociodemographic, COVID-19-related characteristics, and prevalence of anxiety and depression in the sample (n = 1071)  pandemic (Solomou & Constantinidou, 2020 The findings also showed that being female was associated with greater anxiety and occupational burnout and poorer psychological

| DISCUSS ION
QoL. This is in line with research from other countries (Lai et al., 2020;Luceño-Moreno et al., 2020;Xiao et al., 2020) that have reported greater levels of depression, anxiety, and distress in female compared with male healthcare professionals. Most women were nurses or midwives in our sample. Nurses and midwives (vs nonmedical professionals) reported higher levels of occupational burnout. Recent studies attributed the high levels of burnout in nurses during the COVID-19 pandemic to being involved in the direct care of patients as opposed to non-medical professionals and their increased workload during the pandemic (Lasalvia et al., 2021). Despite these negative outcomes for women, they tended to use approach and support-seeking coping to a greater extent compared with men in this study. It is generally expected that these coping styles would offer protection against stress; however, it is worth interpreting this result within the context of the COVID-19 pandemic. Approach coping involves thinking of ways and taking active steps to deal with the stressor and ameliorate its effects, in other words trying to take control of an otherwise uncontrollable situation like the pandemic.
A recent qualitative study found that a key theme in the experience of healthcare professionals during the COVID-19 pandemic was the increased workload along with feelings of losing control (Eftekhar Ardebili et al., 2020). Support-seeking involves seeking support from one's environment and venting of negative emotions, all of which would be constituted difficult during the pandemic due to restrictions in contacts, social distancing, and quarantine measures. This is especially relevant for female healthcare professionals in Cyprus who have an active role within their family environment, often caring for their families. Thus, as suggested by previous studies, seeking quality support by healthcare professionals during the pandemic is hindered by avoiding direct contact due to the need to protect their social circle (Muller et al., 2020). Further qualitative studies may help clarify these underlying coping processes.
Another key finding of the study was that profession and work setting was associated with psychological outcomes. Overall, our findings suggested that medical professionals (vs non-medical), those working in inpatient and ICU settings (vs public health), and those in the frontline (vs non-frontline) experienced greater psychological impact especially in relation to occupational burnout.
Furthermore, nurses and midwives were at a higher risk for depression. These findings are in accordance with a recent meta-analysis on the psychological impact of COVID-19 in healthcare professionals and the role of work setting (Pappa et al., 2020). Inpatient wards and ICUs are at the centre of this pandemic hence increased workload and subsequent burnout were expected. In these units, medical professionals provide direct care to patients, experience the toll of the pandemic, and take important life-death decisions. Our study further revealed that organisational weaknesses can negatively influence healthcare professionals. Specifically, the perception of inadequate workplace preparation was associated with multiple psychological outcomes, including worse QoL outcomes and greater anxiety, depression, and occupational burnout. Our assessment of workplace preparation consisted of a single item; thus, it is not possible to pinpoint the specific organisational weaknesses responsible.
Past research has suggested that lack of personal protective equipment, insufficient training in pandemic response, limited ICU beds and respirators, long shifts, and lack of support are some organisational failings that contributed to worse psychological outcomes in healthcare professionals during the COVID-19 pandemic (Paiano et al., 2020;Suryavanshi et al., 2020).
Our findings also showed that avoidance coping was associated with worse psychological outcomes. Avoidance coping represents strategies including denial and giving up, which have been deemed maladaptive both in the current (Babore et al., 2020) and the SARS epidemic (Wong et al., 2005). The impact of work setting and perceptions of organisational weaknesses on psychosocial outcomes discussed above may be further elucidated by the differences observed in the use of avoidance coping. Interestingly, professionals working in inpatient and ICU settings and those who perceived inadequate workplace preparation used avoidance coping in greater extent than those working in primary settings and who perceived adequate preparation, respectively. It appears that when faced with deficiencies in terms of preparation that are out of their control, professionals working in frontline units, resort in a rather disengaged attitude, which consequently has a negative impact on their psychological health. This is in line with the goodness of fit hypothesis proposed by Lazarus and Folkman (Lazarus & Folkman, 1984), which assumes that emotion-focused coping (e.g., denial) is preferred in uncontrollable situations as opposed to problem-focused coping (e.g., problem-solving) in controllable situations.
In the present study, two of the most common and strongest predictors of poorer QoL were depression and occupational burnout, which is in accordance with previous studies (Stojanov et al., 2020;Suryavanshi et al., 2020). Depression and occupational burnout may make healthcare professionals more vulnerable to experiencing poorer QoL in all aspects of their lives, while depressive mood may also encourage a general negative perception in these outcomes (Berlim & Fleck, 2007). Longitudinal studies will be invaluable in deciphering whether QoL challenges in healthcare professionals persist in the longer term, due to depression and occupational burnout during the pandemic.

| Limitations
The cross-sectional nature of the study means there are inherent limitations in maintaining causality in the observed relationships.
While the electronic distribution of the questionnaires and anonymity maximised recruitment and completion rates, they limited the collection of longitudinal data on the progression of psychosocial challenges in the same sample. Participants were recruited through multiple means. This, along with the anonymous participation, may have inadvertently led to the possibility of duplicate entries by some participants. This limitation was partially addressed by following a complete case analysis, which helped exclude duplicate cases where the survey was incomplete in the first attempt.

| Generalisability and transferability
Much

| CON CLUS IONS
The findings of the present study highlight the psychosocial consequences of the COVID-19 pandemic in healthcare professionals with observed notable levels of anxiety and depression. We have identified several risk factors for poor QoL, anxiety, depression, and occupational burnout, including female gender, medical profession (vs non-medical), frontline units (inpatient wards, ICUs), inadequate workplace preparation to deal with the pandemic, and engaging in avoidance coping. The presence of depression and anxiety further impedes QoL in healthcare professionals. These findings can inform immediate psychological interventions, public health initiatives and future protocols to help healthcare professionals cope with the psychosocial burden of COVID-19 and similar health crises in the future.

CO N FLI C T S O F I NTE R E S T
None declared.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.