Personal, professional, and psychological impact of the COVID-19 pandemic on hospital workers: A cross-sectional survey

Objectives We aimed to evaluate the personal, professional, and psychological impact of the COVID-19 pandemic on hospital workers and their perceptions about mitigating strategies. Design Cross-sectional web-based survey consisting of (1) a survey of the personal and professional impact of the COVID-19 pandemic and potential mitigation strategies, and (2) two validated psychological instruments (Kessler Psychological Distress Scale [K10] and Impact of Events Scale Revised [IES-R]). Regression analyses were conducted to identify the predictors of workplace stress, psychological distress, and post-traumatic stress. Setting and participants Hospital workers employed at 4 teaching and 8 non-teaching hospitals in Ontario, Canada during the COVID-19 pandemic. Results Among 1875 respondents (84% female, 49% frontline workers), 72% feared falling ill, 64% felt their job placed them at great risk of COVID-19 exposure, and 48% felt little control over the risk of infection. Respondents perceived that others avoided them (61%), reported increased workplace stress (80%), workload (66%) and responsibilities (59%), and 44% considered leaving their job. The psychological questionnaires revealed that 25% had at least some psychological distress on the K10, 50% had IES-R scores suggesting clinical concern for post-traumatic stress, and 38% fulfilled criteria for at least one psychological diagnosis. Female gender and feeling at increased risk due to PPE predicted all adverse psychological outcomes. Respondents favoured clear hospital communication (59%), knowing their voice is heard (55%), expressions of appreciation from leadership (55%), having COVID-19 protocols (52%), and food and beverages provided by the hospital (50%). Conclusions Hospital work during the COVID-19 pandemic has had important personal, professional, and psychological impacts. Respondents identified opportunities to better address information, training, and support needs.


Introduction
The COVID-19 pandemic has compelled frontline healthcare workers (HCWs) to risk their personal safety in providing patient care. Such challenging circumstances have adverse effects on frontline HCWs. In the aftermath of the 2003 severe acute respiratory syndrome (SARS) pandemic, HCWs experienced symptoms of depression, anxiety, and chronic stress [1][2][3][4][5] that persisted for years [3].
The COVID-19 pandemic presents unique challenges to HCWs [6][7][8][9][10]. Concerns about personal and family safety are compounded by worries about a surge of patients, depletion of hospital resources (e.g., ventilators, personal protective equipment [PPE]), and rapidly-changing direction from various levels of authority (e.g., PPE recommendations). In contrast to prior public health emergencies (i.e., SARS), our current digital age provides a wealth of ondemand, unverified information that predisposes HCWs to cognitive fatigue. Lastly, physical distancing, while an effective strategy to control the spread of this disease, may lead to personal isolation and a loss of support systems that are vital for HCWs' psychological wellbeing.
Studies conducted in the aftermath of the SARS outbreak reported that being in the nursing profession [2], being a frontline worker [4,5], female gender [5], relationship status [2], and living with a child or children [4] were predictive of increased psychological burden among HCWs. Similarly, emerging evidence from studies during the COVID-19 pandemic suggest that younger age [9,11], working in a community hospital [7,9,10,12], and knowing someone with COVID-19 [6] are additional predictors of increased psychological burden among HCWs.
HCWs are fundamental to the functioning of the increasingly stressed health care system and are the most limited resource in many jurisdictions [13,14]. There is an urgent need to determine the scope of the pandemic's impact on the healthcare workforce and identify and implement public health mitigation strategies [15]. The purpose of this study is to characterize the personal, professional, and psychological impact of the COVID-19 pandemic among teaching and non-teaching hospital workers.

Methods
We conducted a web-based survey of workers, including nurses, physicians, other healthcare professionals, as well as administrative, research, and other hospital staff, at 4 teaching and 8 non-teaching hospitals across 2 regions in Ontario (Toronto and Southwest Ontario [SWO]). The checklist for reporting results of internet e-surveys (CHERRIES) was used for a more complete description of the survey methodology [S1 Appendix] [16]. The study was approved by the Sinai Health (20-0089-E) and Western University (#115850) Research Ethics Boards; and consent was implied by survey completion.

Survey instruments
We adapted a survey previously used during the SARS pandemic [2,5,8]. Using formal survey development methodology [17], the research team iteratively refined the existing instrument and engaged representatives from stakeholder groups (nurses, physicians (including a psychiatrist), health disciplines professionals, and researchers with pandemic expertise) for pre-testing to ensure that questions addressed concerns specific to hospital staff, and were likely to yield information pertinent to the study objective. The survey was available only in English. Responses consisted primarily of attitude statements scored on a six-point Likert scale ranging from (1) Strongly Disagree to (6) Strongly Agree, with an option for free-text responses to several survey items [S2 Appendix].

Sample size calculation
We derived a minimum sample size estimate of 346 using standard survey sample size calculation that incorporates population size, confidence level of 95% and confidence interval of 5%. We aimed to collect a minimum of 1500 complete responses to enable subgroup analyses.

Settings & survey administration
The survey was distributed across 12 hospitals in Ontario over several weeks in July and September 2020, using one of two secure, web-based platforms: NoviSurvey for Toronto sites and REDCap 1 for Southwest Ontario sites. Participants were invited via email by local hospital leadership to complete the survey, with 2 to 3 reminders over several weeks. In addition, the survey invitation and link were posted on the hospital COVID-19 research page for Toronto sites. Survey announcements and email invitations are included in S3 Appendix.

Data analysis
We summarized responses using descriptive statistics: proportions, means and standard deviation (SD), and medians and interquartile ranges (IQR), as appropriate. For Likert-scale questions, we summarized responses according to the proportion of respondents that agreed (either strongly agree, agree, or unsure but probably agree) with each item.
We reported the K10 total score, and depression and anxiety subscores using descriptive statistics [S1 Table]. We performed independent-samples Kruskal-Wallis Test to compare nursing professionals, physicians, and other hospital staff on the K10 psychological distress, depression subscore, anxiety subscore, and IES-R score.
After conducting regression diagnostics (assumption testing), we performed ordinal regression analysis to identify predictors of increased workplace stress (strongly agree, agree, or unsure but probably agree with the statement: "I have felt more stressed at work") and linear regression analysis to identify predictors of K10 total score, depression subscore, and anxiety subscore, and IES-R score. For all outcomes, we selected 11 predictor variables that have shown an association with psychological symptom burden during the SARS outbreak [4,5,8] and the COVID-19 pandemic [9,10]: five predictors related to professional activities (working in a teaching vs. non-teaching hospital, nursing vs. other professions, being a frontline worker, years of healthcare experience, and feeling at increased personal risk due to PPE shortage or inadequate PPE training), and six related to demographic characteristics (age group, gender, high-risk health condition, relationship status, living with one or more children, and knowing someone who contracted COVID-19) [S2 Table]. All variables were entered into the initial regression model and purposefully selected according to the approach described by Bursac and colleagues (2008) [26]. We retained those variables that yielded an association with the outcome variables with a p-value < 0.1 in the final regression model [S2 Table]. We performed all statistical analyses using Statistical Package for Social Sciences Version 25.0 (IBM Corp, 2017; Armonk, NY, USA).
Two investigators (KH and DL) performed qualitative analysis of all open-ended responses using thematic content analysis methodology [27]. First, they coded each open-ended response independently and in duplicate. Then, in a series of coding meetings, they generated themes and subthemes related to the personal and professional impact of the COVID-19 pandemic on HCWs.

Results
Between July and September 2020, 1875 individuals (84% female) completed the survey and were included in the analysis. The sample consisted of nurses (n = 623, 33%), physicians (n = 168, 9%), other health discipline professionals (n = 441, 24%), among other hospital workers (n = 643, 34%). Among these, 923 (49%) were frontline workers (reported caring for patients with suspected or confirmed COVID-19). Table 1 presents respondents' professional and demographic characteristics. Overall, 72% reported being afraid of falling ill and 64% felt that their job placed them 'at a great risk of exposure' to COVID-19. Almost half (48%) reported feeling that they had little control over whether they got infected, 48% reported being preoccupied with their own symptoms, and 40% found it hard to feel reassured of their health. More than one-third of respondents were afraid to tell their family about their professional exposure risk.

Perception of risk
A notable proportion felt that their chance of dying from COVID-19 in the next year was higher than dying from cancer (38%) or a traffic accident (29%), and 17% felt their chances of survival were poor if they contracted COVID-19. Despite these fears, 69% accepted the risk of contracting COVID-19 as part of their job.
Regarding perceived stigmatization due to their profession, 61% (981/1612) felt that people avoided them, 39% (628/1594) felt that people avoided their family members, and 35% (557/ 1615) have avoided telling people about the nature of their job.
More than half (949/1720, 55%) expressed confidence that their employer would look after their medical needs if they were to fall ill with COVID-19, and a similar proportion (899/1648, 55%) felt appreciated by their hospital. In comparison, 75% (1209/1624) reported feeling appreciated by society in general. Table 1 summarizes the demographic characteristics of the 962 respondents in the Southwest Ontario survey who completed the K10 and IES-R scales. Overall, 56% (480/861) had more than mild symptoms of psychological distress or post-traumatic stress, and 38% (317/839) had concern for at least one psychological diagnosis based on the two instruments.

Predictors of workplace stress and psychological symptoms on IES-R and K10
There were no violations of the assumption testing on ordinal and regression analyses. On multivariable ordinal regression, predictors of increased workplace stress (responding strongly agree or agree to the question: "I have felt more stressed at work") included female gender, having a high-risk health condition, younger age, personally knowing someone who contracted COVID-19, working in a non-teaching (relative to a teaching) hospital, and feeling increased personal risk due to PPE shortage or inadequate training ( Table 2).
On multivariable linear regression, predictors of higher psychological distress (K10 total score) included younger age, female gender, having a high-risk health condition, personally knowing someone who contracted COVID-19, and feeling increased personal risk due to PPE shortage or inadequate training. Table 3 shows the final multivariable regression models for K10 psychological distress, as well as depression and anxiety subscores. Predictors of higher symptoms of post-traumatic stress on the IES-R included female gender, having a high-risk health condition, and feeling increased personal risk due to PPE shortage or inadequate training.

Supportive strategies
When asked to select strategies that would help them cope with the COVID-19 pandemic, the most frequently endorsed strategies included: clear and unambiguous communication from their hospital (59%), knowing that their voice is heard and important (55%), expressions of appreciation from hospital leadership (55%), having COVID-19 protocols and procedures (52%), and food and beverages provided by the hospital (50%). Fig 5 presents the proportion of respondents that selected each supportive strategy among all participants and among frontline workers only (those who reported caring for patients with suspected or confirmed COVID-19).  Table].

Discussion
In this cross-sectional survey of hospital workers at 4 teaching and 8 non-teaching hospitals, we found that the COVID-19 pandemic has had important personal impact (e.g., fears and anxieties about exposure, falling ill, exposing others, stigmatization), professional impact (e.g., increased workload, workplace stress, expanded responsibilities), and psychological sequelae. One-quarter of respondents reported at least mild psychological distress on the K10 and nearly half had at least some symptoms of post-traumatic stress on the IES-R; 38% had scores that raise concern for at least one psychological diagnosis.
This study adds important information to the current paucity of data on the personal and professional impact of the COVID-19 pandemic on hospital workers. Overall, our findings are similar to those found in the aftermath of the SARS pandemic [2]. We found that a hospital staff continued their work in the pandemic setting despite believing that their chances of survival from COVID-19 were poor and that their chances of dying from COVID-19 within the next year are higher than that associated with a traffic collision or cancer. More than half of respondents reported stigmatization because of their work at a hospital, and over one-third had avoided telling people about the nature of their job. These rates are similar to a Singaporebased survey during the SARS pandemic, in which 49% HCWs reported that people had avoided them and 31% perceived that people had avoided their family members [2]. The most frequently reported personal coping strategies involved supportive relationships (i.e., talking to friends, family, and colleagues), informational (i.e., learning as much as I can about COVID-19), or attitudinal (i.e., accepting the inherent risk). Nevertheless, many respondents reported reluctance in telling their families about their exposure risk, which could adversely impact their coping. Of note, one-quarter of respondents endorsed use of 'alcohol, marijuana, or other recreational drugs' to cope with the stress of the pandemic and 44% contemplated leaving their job. These findings highlight the urgent need to identify HCWs at high risk of adverse outcomes, and to provide supportive strategies and ensure access to psychological counselling during and after the pandemic.
Similar to previous studies, nursing professional had more psychological distress than physicians [2,7,8]. The increased burden of psychological distress relative to physicians was also seen in other non-physician hospital workers (Fig 3; S3 Table), highlighting the need to ensure that support strategies implemented at the regional and hospital levels should be inclusive of nursing professionals and other health disciplines professionals. Other personal and professional characteristics have emerged as independent predictors of higher psychological symptoms during the COVID-19 pandemic: younger age [9,11], female gender [7,9,10,29], being a frontline worker [7,12], and working in a community hospital [7,9,10,12]. In this study, younger age and female gender were associated with higher workplace stress, psychological distress, as well as depression and anxiety symptoms, and female gender was also associated with higher post-traumatic stress symptoms. We also identified other factors associated with high workplace stress and psychological symptoms: having a high-risk health condition, personally knowing someone who contracted COVID-19, and feeling increased personal risk due to PPE shortage or inadequate PPE training. The latter finding highlights the importance of PPE supply and training in HCWs personal safety and perception of risk, and potentially mitigating the adverse effects of the pandemic.
Emerging evidence have shown that job insecurity and employee burnout during the COVID-19 pandemic influence customer orientation and workplace motivation, with broad implications for human resource management in sectors outside of healthcare [30]. Although job insecurity has not been a concern for healthcare workers as it has been for other workers during the COVID-19 pandemic, HCW burnout is certainly likely to influence workplace motivation and therefore patient care in the healthcare setting. Similar to non-healthcare settings, various organizational changes may mitigate the adverse effects of the COVID-19 pandemic on HCWs and the patient care they provide. In the present study involving employees in the healthcare sector, we identified several opportunities to create a supportive work environment that protects the wellbeing of hospital staff and fosters improved relationships in the workplace. The most frequently favoured supportive strategies were cost-free and primarily associated with the development of a more supportive culture by hospital leadership. These included clear and unambiguous communication from the hospital, HCWs knowing that their voice is heard and important, expressions of appreciation and gratitude from hospital leadership, and having COVID-19 protocols and procedures in place. These findings were also supported by the open-text responses, in which respondents expressed a great need to feel heard, understood, and appreciated by hospital management. Future studies should evaluate the impact of various supportive strategies in mitigating the adverse impact of pandemics on hospital workers.
This study has several limitations. We distributed the survey at a single timepoint and findings may not reflect evolving perspectives of workers as the pandemic persisted. It is not possible to determine an accurate response rate given that the survey was distributed through hospital-wide emails to all staff. More than 80% of respondents were female; while this is consistent with the demographics of the professions most highly represented in this survey (i.e., nursing professionals), the overall findings may best reflect the perspectives of women. Finally, hospital-wide distribution of the survey and inclusion of hospital staff with a broad range of professional characteristics disallows any conclusions about specific subgroups of hospital workers. This study has several strengths. First, we adapted a survey instrument that was used to evaluate the impact of the SARS pandemic on healthcare workers [5,8], and administered two validated instruments evaluating psychological distress and symptoms of post-traumatic stress. We included a large sample of respondents representing a broad range of professions and practice settings (teaching and non-teaching) to ensure that the findings are generalizable and to inform supportive strategies applicable to a broad range of professions.

Conclusion
In this cross-sectional survey of staff at 12 teaching and non-teaching hospitals, we found that the first wave of the COVID-19 pandemic had important personal, professional, and psychological effects on hospital workers. We identified several low-cost opportunities for healthcare systems and hospitals to support and address the needs of hospital workers during pandemics, including clear and unambiguous communication with staff, recognizing that the voice of HCWs is important, expressions of gratitude and appreciation by hospital leadership, and having COVID-19 protocols and procedures in place.