1. Background and Hypothesis

COVID-19 is an infectious disease caused by SARS-CoV-2. The first cases were reported in Wuhan (China), then the infection spread all over the world. The World Health Organization (WHO) on March 11, 2020 declared the novel coronavirus outbreak a global pandemic. The virus has already had a direct impact on the physical health of millions of people, also it is supposed to be a mental health threat of great magnitude [1]; in fact, not only the pandemic, but restrictive measures such as the lockdowns have dramatically affected people’s everyday life: in particular, the rapid spread of the virus has reduced the chances of social interaction including transformed them in potentially dangerous situations [2]. Several studies have found an association between the COVID-19 pandemic and psychiatric symptoms, such as distress, anxiety, fear of infection, depression, and insomnia both in the general population and among vulnerable populations including people with pre-existing psychiatric disorders [1]. Gloster et al. [3] assessed 9,565 participants from 78 countries to examine the impact of the pandemic and resultant governmental restrictive measures on mental health. During the peak of stay-at-home orders, the pandemic was experienced as, at least, moderately stressful for almost the entire sample, while 11% reported the highest levels of stress. Symptoms of depression were also present at a high level, and 33% reporting high levels of boredom, and nearly 50% indicating they wasted a lot of time. Consistent with symptoms of stress and depression, 10% of participants were psychologically languishing [3]. These results suggest that there is a subgroup of people particularly prone to COVID 19 mental health consequences, and that nearly 50% of the assessed population reported at least a moderate reduction of mental well-being [3].

Healthcare workers have been professionally overloaded, trying to manage the psychosocial impact of the pandemic and suffering its effects in person. Previous studies demonstrated an increase of distress and burnout among healthcare workers during epidemics: surprisingly, in many cases they were affected by severe Post-Traumatic Stress (PTS) symptoms, as shown by Wu and colleagues [4]. In the study a sample of 549 hospital employees was assessed during 2003 SARS epidemic and during the 3-year period following the outbreak. Nearly 40% of hospital employees suffered from elevated PTS symptoms three years after the SARS outbreak. They also found that exposure to the SARS outbreak at work, being quarantined, and the death or illness of a relative or friend from SARS, each contributed independently to PTS symptom levels [4].

A recent study [5] evaluated distress and burnout due to the COVID19 pandemic among mental health workers in the Lombardy region in Italy. Main findings showed a mild stress response during the pandemic, with 6.6% of the sample experiencing moderate to severe levels of depression and 11.6% showing moderate to severe anxiety. These results are similar to a recent Chinese study which revealed a high prevalence for mental health symptoms among healthcare workers treating patients with COVID-19 in China. Overall, participants reported mainly symptoms of distress (71.5%) [6].

The aim of the present study was to evaluate disease perceptions, distress and perceived support among healthcare workers in Italy at three different times, which correspond to the main phases of the pandemic management in Italy: T0, the first peak of the infection rate and the stay-at-home order, from March 2020 to June 2020; T1, reduction of the infection rate and the reopening, from June 2020 to October 2020; T2, the second wave of infections with a new progressive closure, from October 2020 to December 2020. We collected data from different healthcare professionals to differentiate between those who were on the frontline and those on the second line, and how the distress was perceived, according to the level of direct and indirect exposure.

We hypothesized that: 1) levels of distress would be higher in operators exposed to COVID-19 (i.e. direct and family contagion) and in frontline operators compared to those on the second line; 2) levels of distress would be inversely correlated to the perceived degree of support obtained by the team and managers; 3) levels would be considerably higher at T0 and T2 compared to T1.

2. Methods

The open survey was designed to target Italian health professionals through social media (Facebook and Instagram). Questionnaires were distributed electronically over a 2-week period from 18^th November to 2^nd Dec 2020. The survey was conducted in different regions across Italy. The final convenience sample included 514 healthcare workers, recruited via social media. Before starting the survey, participants had to give their informed consent to continue. Informed consent included the purpose of the study, those responsible for it and information on the confidentiality of the data, anonymity, and personal data protection. Before completing the survey submission, participants were required to respond to all items. Respondents were able to review and change their answers before submitting the questionnaire. The IP address of the participant computer was used to identify potential duplicate entries from the same user. More entries for the same IP address were never allowed. The completion time for all items was approximately 10-15 min. No incentive was offered for participation. Demographic data including sex (male or female), age, geographic location, marital status, number of cohabitants, occupation (e.g. physician, nurse, technician, or other healthcare professionals), work location (e.g. hospitals, outpatient services), medical discipline (e.g. internal medicine, general surgery, intensive care unit, imaging, etc.) were self-reported by responders. The study was conducted according to the principle of the Declaration of Helsinki.

We used four validated questionnaires: the BIPQ (Brief Illness Perception Questionnaire) [7], the PSS-10 (Perceived Stress Scale-10) [8], the BAI (Beck Anxiety Inventory) [9] and the BDI (Beck Depression Inventory) [10]. We also used a survey (13 items) at T0, T1 and T2 to evaluate exposure, perception, quality of life, and burnout. In addition, we asked to evaluate the perception of the support obtained by the team and the manager (2 items) which were rated on a visual analogue scale, ranging from 0 (no support at all) to 10 (best support). BIPQ is a nine-item scale designed to rapidly assess the cognitive and emotional representations of illness. The PSS-10 is the most widely used psychological instrument for measuring the perception of stress. It is a measure of the degree to which situations in one’s life are appraised as stressful. The ten items were designed to tap how unpredictable, uncontrollable, and overloaded respondents find their lives. The scale also includes some direct queries about current levels of experienced stress. The BAI is a 21-item brief measure of anxiety with a focus on somatic symptoms of anxiety that was developed as a measure for discriminating between anxiety and depression. The BDI is a 21-item multiple-choice self-report inventory, one of the most widely used psychometric tests for measuring the severity of depression. Items of each questionnaire were presented in the order required by the protocol and not randomized.

3. Statistical Analysis

Descriptive statistics were presented for all variables. As all the included variables were not normally distributed, nonparametric tests were applied. A generalized linear model was constructed in order to evaluate the temporal change in PSS-10, BAI and BDI scores. McNemar’s test was used to evaluate changes between the three time points in frequency use of psychotropic medications. Bonferroni’s correction for multiple comparisons was applied. A two-tailed p-value<0.05 was regarded as statistically significant. All calculations were performed using Stata 16 (Stata Corp, College Station, Texas 77845 USA).

4. Results

Results are presented according to the Checklist for Reporting Results of Internet E-Surveys (CHERRIES) (see supplementary material) [4]. General characteristics of the study sample are depicted in Table 1. Overall, 514 people completed the online survey.

Healthcare professionals exposed to COVID 19 patients (n=318) showed higher PSS scores (21.35 ± 7.46 vs. 18.80 ± 7.14, U=25070.5, p<0.001), BAI scores (13.75 ± 11.46 vs. 9.91 ± 9.43, U=24663, p<0.001) and BDI scores (10.74 ± 10.11 vs. 8.29 ± 8.97, U=26067, p=0.002) at T0 compared to healthcare professionals not exposed (n=196). The same effect was observed at T1 for PSS scores (18.10 ± 6.99 vs. 15.45 ± 6.41, U=24652.5, p<0.001), BAI scores (8.91 ± 9.19

5. 5.86 ± 7.23, U=25393.5, p<0.001) and BDI scores (8.49 ± 8.82 vs. 5.37 ± 7.28, U=25998.5, p<0.001) in healthcare professionals dealing with COVID 19 patients (n=221). At T2, again the same effect was observed for PSS scores (21.53 ± 7.53 vs. 19.27 ± 7.16, U=23223.5, p=0.001), BAI scores (13.09 ± 10.29 vs. 9.54 ± 8.77, U=22302, p<0.001) and BDI scores (11.76 ± 9.25 vs. 8.59 ± 8.68, U=21696.5, p<0.001) in healthcare professionals dealing with COVID 19 patients (n=352). Having a family member infected by COVID 19 (n=71) determined higher PSS score (22.08 ± 7.44 vs. 20.11 ± 7.41, U=13343, p=0.04), BAI scores (16.85 ± 12.72 vs. 11.56 ± 10.39, U=11773, p=0.001) and BDI scores (12.61 ± 11.04 vs. 9.35 ± 9.47, U=12732, p=0.01) at T0. The same effect was not observed at T1, however in a reduced sample (n=16) as well as at T2 (n=88). Having experienced a family death for COVID 19 did not significantly increase PSS, BAI and BDI scores at T0 (n=16), T1 (n=3) and T2 (n=6). Healthcare professionals who were infected by COVID 19 or developed COVID-like symptoms (n=85) reported higher PSS score (24.35 ± 6.68 vs. 19.59 ± 7.34, U=11320.5, p<0.001), BAI scores (19.84 ± 12.12 vs. 10.79 ± 9.98, U=9832, p<0.001) and BDI scores (15.59 ± 11.38 vs. 8.66 ± 8.98, U=10791.5, p<0.001) at T0. The same effect was observed at T1 (n=21) for PSS (19.71 ± 6.06 vs. 16.46 ± 6.79, U=3684, p=0.02) and BAI scores (11.52 ± 9.99 vs. 6.99 ± 8.14, U=3670.5, p=0.02), but not for BDI score (10.71 ± 10.24 vs. 6.54 ± 7.99, U=3967.5, p=0.07). At T2, healthcare professionals who were infected by COVID 19 or developed COVID-like symptoms (n=67) reported higher PSS score (22.51 ± 6.29 vs. 20.57 ± 7.62, U=12569, p=0.03), BAI scores (15.88 ± 8.90 vs. 11.39 ± 9.99, U=9979.5, p<0.001) and BDI scores (13.24 ± 9.62 vs. 10.39 ± 9.07, U=11832.5, p=0.006).

Table 1: General characteristics of study participants.

Correlations between variables are reported in Table 2. Overall, levels of perceived support from the medical directors and the team correlated positively at all-time points (all p<0.05). Additionally, levels of perceived support were negatively related to PSS 10 scores at all-time points (all p<0.05).

Table 2: Correlation coefficients between study variables.

A GLM analysis reported a significant difference in PSS 10 scores across the three time points (F=150.89, p<0.001): specifically, scores at T0 were higher than scores at T1 (Mean difference 3.79 CI 95% 3.27-4.31, p<0.001) but not at T2 (Mean difference -0.44 CI 95% -0.97-0.08, p=0.1). Additionally, scored at T1 were significantly lower compared to T2 (Mean difference -4.23 CI 95% -4.76/-3.70, p<0.001).

A GLM analysis reported a significant difference in BAI scores across the three time points (F=152.44 p<0.001): specifically, scores at T0 were higher than scores at T1 (Mean difference 5.11 CI 95% 4.42-5.81, p<0.001) but not at T2 (Mean difference 0.32 CI 95% -0.40-1.03, p=0.39). Additionally, scored at T1 were significantly lower compared to T2 (Mean difference -4.80 CI 95% -5.43/-4.18, p<0.001).

A GLM analysis reported a significant difference in BDI scores across the three time points (F=117.64 p<0.001): specifically, scores at T0 were higher than scores at T1 (Mean difference 3.09 CI 95% 2.46-3.72, p<0.001) and lower than scores at T2 (Mean difference -0.96 CI 95% -1.66/-0.26, p=0.007). Additionally, scored at T1 were significantly lower compared to T2 (Mean difference -4.05 CI 95% -4.61/-3.50, p<0.001).

Use of anxiolytics was significantly higher in T0 (16.3%) compared to T1 (9.3%) (p<0.001), and in T2 (15%) compared to T1 (9.3%) (p<0.001). No significant difference was observed in use of antidepressants, mood stabilizers or antipsychotics at any time points.

5. Interpretation of the findings

The present study observed that healthcare professionals exposed to COVID 19 presented higher levels of stress, anxiety, and depression. At the same time, being infected by COVID 19 during work exerted a similar effect on level of distress. This is in line with recent evidence that reported that frontline healthcare workers experienced higher distress during pandemics [12,13]. Healthcare workers may experience post-traumatic symptoms as well as depression and anxiety.

On the other hand, we did not observe a significant impact of COVID 19 related deaths in family members on levels of distress in healthcare professionals: this could be partly explained by the luckily low number of COVID 19 deaths in our sample.

Levels of distress were significantly higher during the first wave of the pandemic (March - May 2020) and the second wave (October - November 2020) as compared with the inter-wave period, with no difference between the two waves. During the interval between the two waves, when life was returning to a new normal, distress was significantly lower. During the two waves in Italy, COVID-19 wards were created ex novo or existing wards (i.e., internal medicine, rehabilitation) were transformed in COVID-19 wards. Several healthcare workers were abruptly moved from their previous occupation to attending highly infectious patients, with brief training, while others continued to attend to their usual chores. This could have caused significant distress, which could have been mitigated in the interval between the two waves, when several COVID-19 wards were closed or returned to their original use [14].

Levels of distress experienced by healthcare professionals at each time point were inversely correlated by perceived support from team and medical managers. This finding could move the focus from a mere description of psychological consequences of COVID-19 on health workers to a more proactive stance: as it is impossible to eliminate pandemic stress, every strategy to increase resilience or reduce vulnerability to burnout could be useful [15,16]. Management and organizational support may foster positive feelings about work and a better ability to cope with work stress [17]. Medical managers should overview work schedules in order to distribute work shifts adequately and allow for an adequate number of sleep hours [18]. Moreover, being part of a group and feeling cooperation and trust among team members are generally associated with well-being through shared experiences [19].

The present survey presents both strengths and limitations. The main strength relies in the wide sample of healthcare professionals reached and by the focus on the role of organizational support to potentially counteract the negative impact of pandemics on health workers. Several limitations of the study should be carefully considered to avoid over interpretation of the findings.

Firstly, the survey was conducted during the second wave of the pandemic and was based on the recall of two different (previous) time-points. Presence of recall bias is therefore a major problem; however, our results are still valid even if we focused on the second wave of the COVID 19 pandemic which happened during the survey. Secondly, most of the sample was composed by medical doctors followed by nurses and therefore study findings may not be generalized to every healthcare professional. Finally, we could not quantify levels of COVID 19 exposures, but we relied on the subjects’ own perception of exposure.

Possible implications

The present survey shed more light on the topic of work-related distress during the COVID 19 pandemic among healthcare professionals. Presence of higher levels of distress among frontline workers calls for actions to improve resilience and prevent potential major consequences (post-traumatic stress disorder, major depression and anxiety) in this professional category.

Ethical standards

The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2000.

References

1. Xiong J, Lipsitz O, Nasri F, et al. Impact of COVID-19 pandemic on mental health in the general population: A systematic review. J Affect Disord 277 (2020): 55-64.
2. Marroquin B, Vine V, Morgan R. Mental health during the COVID-19 pandemic: Effects of stay-at-home policies, social distancing behavior, and social resources. Psychiatry Res 293 (2020):
3. Gloster AT, Lamnisos D, Lubenko J, et al. Impact of COVID-19 pandemic on mental health: An international study. PLoS One 15 (2020): e0244809.
4. Wu P, Fang Y, Guan Z, et al. The psychological impact of the SARS epidemic on hospital employees in China: exposure, risk perception, and altruistic acceptance of risk. Can J Psychiatry 54 (2009): 302-311.
5. Rapisarda F, Vallarino M, Cavallini E, et al. The Early Impact of the Covid-19 Emergency on Mental Health Workers: A Survey in Lombardy, Italy. Int J Environ Res Public Health 17 (2020).
6. Lai J, Ma S, Wang Y, et al. Factors associated with mental health outcomes among health care workers exposed to coronavirus disease 2019. JAMA Netw Open 3 (2020): e203976.
7. Broadbent E, Petrie KJ, Main J, et al. The brief illness perception questionnaire. J Psychosom Res 60 (2006): 631-637.
8. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav 24 (1983): 385-396.
9. Beck AT, Epstein N, Brown G, et al. An inventory for measuring clinical anxiety: psychometric properties. J Consult Clin Psychol 56 (1988): 893-897.
10. Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 4 (1961): 561-571.
11. Eysenbach G. Improving the quality of Web surveys: the Checklist for Reporting Results of Internet E-Surveys (CHERRIES). J Med Internet Res 6 (2004): e34.
12. Alonso J, Vilagut G, Mortier P, et al. Mental health impact of the first wave of COVID-19 pandemic on Spanish healthcare workers: A large cross-sectional survey. Rev Psiquiatr Salud Ment 14 (2020): 90-105.
13. Salazar De Pablo G, Vaquerizo-Serrano J, Catalan A, et al. Impact of coronavirus syndromes on physical and mental health of health care workers: Systematic review and meta-analysis. J Affect Disord 275 (2020): 48-57.
14. Carmassi C, Foghi C, Dell'oste V, et al. PTSD symptoms in healthcare workers facing the three coronavirus outbreaks: What can we expect after the COVID-19 pandemic. Psychiatry Res 292 (2020): 113312.
15. Callahan K, Christman G, Maltby L. Battling Burnout: Strategies for Promoting Physician Wellness. Adv Pediatr 65 (2018): 1-17.
16. Schreiber M, Cates DS, Formanski S, et al. Maximizing the resilience of healthcare workers in multi-hazard events: lessons from the 2014-2015 ebola response in Africa. Mil Med 184 (2019): 114-120.
17. Maunder RG, Lancee WJ, Mae R, et al. Computer-assisted resilience training to prepare healthcare workers for pandemic influenza: a randomized trial of the optimal dose of training. BMC Health Serv Res 10 (2010): 72.
18. Dewey C, Hingle S, Goelz E, et al. Supporting clinicians during the COVID-19 pandemic. Ann Intern Med 172 (2020): 752-753.
19. Mahmoud NN, Rothenberger D. From Burnout to Well-Being: A Focus on Resilience. Clin Colon Rectal Surg 32 (2019): 415-423.