The Impact of Epidemics and Pandemics on the Mental Health of Healthcare Workers: A Systematic Review

1The frequency of disease outbreaks has increased over the past century due to population growth, the increased interconnectedness of the world, microbial adaptation and change, economic development, changes in land use, and climate change. Emerging infectious diseases that have caused epidemics over the past two decades include the severe acute respiratory syndrome (SARS) in 2003, Influenza A virus subtype H1N1 in 2009, Middle East respiratory syndrome coronavirus (MERS-COV) in 2012, Ebola Virus Disease (EVD) in 2014, the influenza A virus subtype H7N9, and the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in December 2019, which has resulted in the coronavirus disease 2019 (COVID-19) pandemic.

2Disease outbreaks cause an unexpected increase in morbidity and mortality, which in turn cause an increased demand on healthcare facilities. The rapid increase in patient populations drastically reduces the healthcare worker (HCW) to patient ratio thus increasing workload. HCWs suffer from both physical and mental fatigue because their working hours are increased and they may be asked to work more night shifts; thus, they do not have enough time to sleep, rest, and recuperate. As they work in the frontline, diagnosing, managing, and caring for sick patients, they experience a variety of mental health symptoms which may also persist after the epidemic has ended.

3The massive influx of patients overwhelms the capacity of healthcare systems, giving rise to ethical dilemmas around the distribution of essential healthcare and medical supplies. HCWs constantly have to make “life or death” decisions, such as which patients to admit or not admit into intensive care and when to withdraw life support. Due to the increased numbers of people dying, HCWs repeatedly break bad news, sometimes in ways they are not used to, including over the phone, thus making breaking bad news more distressing. The news of continuously rising numbers of confirmed cases and deaths is emotionally overwhelming.

4The shortage in supplies of personal protective equipment (PPE) may increase the risk and fear of contagion. During this time, HCWs continuously live with anxiety and fear of contracting the disease more so when a colleague becomes infected or dies. They fear transmitting the infection to their families as well as experiencing stigma and discrimination from their communities due to transmission fears. This type of stigmatization may even escalate to harassment, being denied access to public transport, physical violence, and eviction from their homes by landlords. Social ostracization aggravates the occupational stress that HCWs are already facing as they battle the disease outbreak.

5Overall, these negative psychological factors do not only affect the HCWs themselves, but also reduce their effectiveness in fighting epidemics, therefore indirectly affecting the whole population at large. This systematic review aims to synthesize the available evidence on the impact of epidemics and pandemics on the mental health of HCWs which will guide and inform best practice policies for psychological supports and mental health interventions for HCWs. Even though similar systematic reviews have been conducted recently, these were either specific to a single pandemic (i.e., SARS) or Covid-19, or with a small number of studies, and included only one study from low- and middle-income countries.

6This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols guidelines (PRISMA). However, the review is descriptive in nature, and the data extracted from the selected studies were summarized but not statistically combined owing to methodological heterogeneity. The study protocol was pre-registered with the National Institute for Health Research international prospective register of scientific reviews (PROSPERO, CRD42020186331).

7A comprehensive systematic search of the most common databases—PubMed, PsycInfo, and PsycArticles, was conducted from inception to June-end 2020. Furthermore, the reference list of the retrieved articles and systematic reviews of similar topics were also examined to verify whether any potential studies had been left out. The author (O.C.C.) conducted the initial literature search. The full search strategy is available in Appendix A. Combinations of the following terms were used for the search:

8Category 1: Population (healthcare professional, healthcare workers, physician, doctor, nurse)

9Category 2: Exposure (epidemic, pandemic, SARS, MERS, Ebola, H1N1, H7N9, COVID 19). The terms epidemic and pandemic were defined according to the World Health Organization definitions. “An epidemic is the occurrence in a community or region of cases of an illness, specific health-related behavior, or other health-related events clearly in excess of normal expectancy”. “A pandemic is the worldwide spread of a new disease”.

10Category 3: Outcomes (mental health, mental disorder, psychological, depression, anxiety, stress, burden, insomnia, sleep disturbance, burnout, fear, stigma, discrimination).

11Search results citations were downloaded to Endnote reference management software version X9 and duplicates were removed. The author (O.C.C.) performed the initial screening of titles and abstracts for relevance.

12The following criteria were applied for papers to be included in the review:Studies reporting the impact of a pandemic/epidemic on mental health outcomes of health care workers.Cross-sectional, case–control, and cohort studies. Intervention studies were considered for inclusion only when they had sufficient details about the baseline mental health outcomes.Studies were selected if data was from an original studyStudies had to be published in a peer reviewed journal.Only English language studies were included.No restrictions were placed on the publication date.There was no limit on the geographical location of studies.

13Studies reporting the impact of a pandemic/epidemic on mental health outcomes of health care workers.

14Cross-sectional, case–control, and cohort studies. Intervention studies were considered for inclusion only when they had sufficient details about the baseline mental health outcomes.

15Studies were selected if data was from an original study

16Studies had to be published in a peer reviewed journal.

17Only English language studies were included.

18No restrictions were placed on the publication date.

19There was no limit on the geographical location of studies.

20Preprints, study protocols, and conference abstracts/proceedings were excluded.

21The author (O.C.C.) checked the relevant studies for eligibility and extracted data from the eligible studies onto a standard Microsoft Excel data extraction form. A second reviewer (A.S.) independently verified the eligibility of the included studies. Any discrepancies were resolved by discussion. Full text articles for eligible studies were obtained. The data extraction form included the author(s) of the study, the publication year, country of study, details about study participants, study settings, study design, outcome measures used, and main findings. Furthermore, information necessary for evaluating study quality was also extracted from the eligible studies. Studies were assessed for methodological quality using the Joanna Briggs Institute (JBI) critical appraisal tools for cross sectional and cohort studies.

22The authors analyzed and synthesized the results using a narrative text approach to summarize and explain the study findings focusing on prevalence of mental health outcomes, and the associated risk and protective factors.

23The three database searches yielded 5716 articles. After removal of 793 duplicates, the titles and abstracts of 4923 articles were screened. Two-hundred-and-thirty potential studies were identified, and the full texts were checked for eligibility. Sixty-eight articles met the inclusion criteria and eight more were identified through searching references of selected papers totaling 76 final studies. Details are provided in the PRISMA flowchart (Figure 1).

24The characteristics of the selected studies are shown in Table 1 and Table 2. Overall, seventy-six papers met the inclusion criteria. Of these, 34 (44%) focused on SARS, 28 (37%) on COVID-19, seven (9%) on MERS, four (5%) on Ebola, 2 (3%) on H1N1, and one (1%) on H7N9. The studies were conducted in different countries: 26 (34%) China; nine (12%) Taiwan; seven (9%) Canada; eight (11%) Hong Kong; seven (9%) Singapore; four (5%) Saudi Arabia; four (5%) Korea; one (1%) each from Germany, Greece, Iran, Italy, Japan, Liberia, Sierra Leonne, Nigeria, Turkey, and USA; and one was conducted in two countries Singapore and India. Most studies were conducted in a hospital setting 71 (95%), three in a general practice setting, and one at a rehabilitation center. Sixty studies (80%) included more than one type of HCW, 12 had only nurses, and three had only doctors/physicians. A higher proportion of studies 71 (93%) studies were cross-sectional and only 5 (7%) were cohort studies.

25A more detailed assessment is available in Table 3 and Table 4. All eligible studies were included in the review, regardless of their quality assessment results. Of the 71 cross-sectional studies, 42 papers (59%) were of very good quality, five papers (7%) were of good quality, 15 papers (21%) were of average quality, and nine papers (13%) were of poor quality. Of the five cohort studies, one paper was of very good quality, two papers were of good quality, one paper had average quality, and one was of poor quality.

26The Impact of Events Scale (IES) and the Perceived Stress Scale (PSS) were the most common instruments used to measure stress. The Generalized Anxiety Disorder (GAD) and the Zung Self-Rating Anxiety Scale (SAS) were frequently used instruments to measure anxiety. Commonly used instruments to measure depression were the Patient Health Questionnaire (PHQ) and the Zung Self-Rating Depression Scale. Insomnia was often measured using the Insomnia Severity Index (ISI) and the Pittsburgh Sleep Quality Index (PSQI). Most studies which measured burnout used the Maslach’s Burnout Inventory.

27Stress was the most commonly measured mental health symptom. Any one of acute stress, distress, or post-traumatic stress symptoms was examined in forty-two studies. The prevalence of stress varied, and it ranged from 5% to 80%. Ten studies identified that nurses experienced more distress compared to doctors. HCWs providing direct care to confirmed cases of SARS and COVID 19 were more likely to be distressed compared to those who did not provide direct care. Moving from a low risk ward to work in a high risk ward, more working time per week, frequent changes in infection control measures and protocols, seeing a colleague getting sick, being intubated or dying increased stress while those who received adequate social support were least likely to have PTSD. Having been in quarantine during the outbreak was associated with high levels of PTSD. Availability of adequate PPE significantly reduced stress.

28Anxiety and fear symptoms were examined in 29 studies. The prevalence of anxiety varied and ranged from 7% to 78% across all virus exposures. Nine studies found that HCWs who had contact with confirmed cases had more anxiety compared to HCWs who had had no contact with confirmed cases. A common cause of anxiety was worrying about transmitting infection to family members. Nurses had higher anxiety scores compared to doctors. Female healthcare workers were more likely to have anxiety compared to males. Three studies from China compared anxiety levels of HCWs in Wuhan to those of HCWs in the outreach or other regions and found that HCWs in Wuhan, which was the epicenter of COVID-19 at that time, had significantly higher anxiety compared to HCWs in other regions of China. Similar results were found in Canada were HCWs in Toronto who had more contact with SARS patients had higher levels of burnout and distress compared to HCWs in Hamilton where they had fewer confirmed cases. Fear and anxiety were significantly increased when a colleague became infected or died. Anxiety and fear of infection were inversely related to availability of hospital resources, HCWs’ resilience and support from family and friends. The increase in working hours during a disease outbreak was directly related to anxiety levels. Lack of knowledge of the virus was also associated with an increase in anxiety.

29Symptoms of depression were examined in 25 studies. The prevalence of depression ranged from 8.9% and 74.2%. Five studies showed that depression was higher in females compared to males. The frontline medical staff working in the respiratory, emergency, ICU, and infectious disease departments were twice more likely to suffer from depression than the non-clinical staff. Nurses working in SARS units were more depressed than nurses in non-SARS units. The HCWs in Wuhan, which was the epicenter of the COVID-19 pandemic, had higher levels of depression compared to HCWs outside Hubei province. Increased working hours were associated with elevated depression and hopelessness. Having a past exposure to traumatic events or pre-existing psychiatric disorder before the epidemic was associated with high levels of depressive symptoms. Those HCWs with a marital status of being single were more likely than married HCWs to have high levels of depressive symptoms. A history of being quarantined was associated with higher levels of depression. Support from family and friends, psychological preparedness, altruistic acceptance, and perceived efficacy of dealing with the pandemic was associated with lower levels of depression.

30Insomnia and sleep quality was assessed in 11 studies. All 11 studies reported substantial sleep problems, ranging from 26% to 45%. Insomnia was independently associated with depression and anxiety. In three studies, insomnia symptoms were higher in frontline HCWs compared to second line workers. Nurses reported more sleep problems compared to other HCWs, and nurses working in SARS units were more likely to have insomnia compared to nurses working in non-SARS units. HCWs in Wuhan reported more insomnia symptoms compared to healthcare workers in other areas out of Hubei province.

31Burnout (emotional exhaustion) was assessed in eight studies, and they all confirmed high levels of burnout in HCWs. HCWs who worked in the frontline or had contact with confirmed cases were more likely to be emotionally exhausted compared to HCWs who were not in the frontline and who had no direct contact with confirmed cases, while one study reported different results in that front-line HCWs had lower levels of burnout compared to other HCWs. The possible explanation given by the researchers for this unexpected trend was front-line HCWs had received timely and accurate information hence they had a higher sense of control of their situation. Two studies showed that HCWs who had spent more time in quarantine had higher levels of burnout. Lower levels of organizational support, job stress and poor hospital resources, were directly related to emotional exhaustion. Burnout was negatively correlated to self-efficacy, resilience and family support. High anxiety scores predicted high levels of burnout.

32Five studies examined stigma and in all studies, HCWs had been stigmatized either by their family or by the community or both. The prevalence of stigma in HCWs ranged from 20% to 49%. HCWs who were working in direct contact with confirmed cases and those who had been quarantined experienced higher levels of stigma. One study which compared psychological morbidity of stigma between general practitioners and Chinese traditional practitioners found that general practitioners had more exposure to SARS patients and suffered more stigma than the Chinese traditional practitioners

33This review showed that epidemics and pandemics have a negative impact on the psychological wellbeing of HCWs by the wide range of mental health symptoms, in particular stress, depression, anxiety, insomnia, fear, stigma, and emotional exhaustion.

34This review identified common factors that increased the risk of mental health symptoms. Frontline HCWs working in high risk environments where they had direct contact with suspected and confirmed cases of SARS and COVID 19 reported more psychological symptoms compared to non-frontline HCWs working in low risk environments. Working in direct contact with infectious patients was associated with higher levels of symptoms of anxiety, stress, insomnia, and depression due to the increased fear of contracting infection, greater concern of infecting family members, stigmatization, and isolation. This might explain why nurses were found to be more stressed, anxious, depressed, and had poorer sleep quality compared to doctors. Most studies explained this to be due to the higher workload that nurses have and the more time they spend in direct contact with patients whilst nursing them. HCWs in the epicenter of a pandemic experienced more psychological distress compared to HCWs in other regions due to the higher exposure to infectious patients. Another occupational risk factor identified was the extent of healthcare experience that a HCW had. HCWs with less work experience were more likely to be stressed compared to HCWs with more years of work experience. Less experienced HCWs have less knowledge, skills, and are less able to self-regulate, thus they get stressed more easily compared to more experienced HCWs who have more knowledge and skills, and are thus more able to adapt.

35Inadequate hospital equipment and the limited supply of personal protective equipment (PPE) were also associated with higher levels of psychological symptoms. Being of female gender was also identified as a risk factor. A history of exposure to other traumatic events before an t outbreak increased the risk of re-occurrence of a psychiatric disorder. Having a high perceived risk of infection and low self-efficacy were also identified as risk factors associated with mental health symptoms. HCWs who were unconfident about beating the outbreak were more depressed and had a poor mental state compared to HCWs who were more confident and resilient. Lack of knowledge of the virus and lack of outbreak management training was associated with low perceived self-efficacy. Constantly changing infection control measures and documentation processes also reduced self-efficacy and caused an increase in stress levels. Having been quarantined was identified as a risk factor of depressive and post-traumatic stress symptoms. This was attributed to the increased fear of dying from the disease. Quarantining was associated with increased levels of fear and stress in HCWs due to the emotional isolation and loneliness experienced during quarantine.

36Despite the limited number of cohort studies compared to cross sectional studies, the cohort studies conducted during the SARS epidemic confirmed the persistence of mental health symptoms up to a year after the pandemic has ended.

37Protective factors identified in this systematic review include adequate information, clear guidelines, training and organizational support, altruistic acceptance of risk,, availability of specialized equipment for treating patients, adequate personal protective equipment, having more years of healthcare experience, adequate time off work, and support from family and friends.

38The strengths of this review are, first, that it identified a large number of studies conducted during and after the epidemics and pandemics that have occurred in the past twenty years, including the current COVID-19 pandemic. Second, results are generalizable as the included studies were from Asia, Europe, Africa, Middle East, and America. Third, most papers included in this review used standardized and previously validated instruments for measuring mental health symptoms. However, a potential limitation is that we only included published articles and excluded gray literature, which might have caused some publication bias. Another limitation is that there were only five cohort studies, 94% of the studies included were cross-sectional which implies that no causal inferences can be drawn. Furthermore, meta-analyses were not undertaken because of the methodological heterogeneity of the studies.

39It is important to conduct more cohort studies to obtain a detailed picture of mental health symptoms at the different points of a disease outbreak, and to understand the long-term mental health impact of a pandemic or epidemic among HCWs.

40The possible role of occupation and exposure on mental health needs to be examined further in future studies. While many studies have reported higher levels of mental health problems among female HCWs, it is still unclear whether gender is a sole influencing factor, or if gender is being confounded by other factors. For instance, most of the female HCWs were nurses, and nurses experience higher mental health problems due to their increased exposure and nature of work. Besides, previous studies have shown that nurses and doctors working in the emergency department and intensive care units are at a higher risk of burnout, depression, and job stress compared to their colleagues working in other hospital departments. Therefore, future studies need to rule out these aspects, while determining the effects of a pandemic or epidemic on mental health.

41Increasing age, and prior chronic medical conditions make a person more susceptible to the effects of a pandemic. Therefore, in future studies, it is important to address the association between these factors and mental health outcome.

42Many studies used online platforms for data collection, and this method is known to increase the risk of sampling and response bias. However, we consider this method as appropriate for the current studies as face-to-face data collection was not possible due to social distancing guidelines.

43As this review identified many protective factors including adequate information about the pandemic, clear guidelines and training, social support, availability of specialized equipment for treating patients, adequate personal protective equipment, adequate time off work, may be provided to the HCWs for reducing adverse mental health outcome.

44This systematic review provides a comprehensive narrative synthesis of the underlying negative impacts of epidemics and pandemics on the mental health of HCWs which include acute stress, post-traumatic stress disorders, severe depression, anxiety, burnout, insomnia, and stigmatization. It is apparent from this review that the current healthcare systems and many governments across the globe need to prioritize mobilizing resources to provide sufficient and necessary psychological support to HCWs during and after epidemics and pandemics.