Did the prevalence of depressive symptoms change during the COVID-19 pandemic? A multilevel analysis on longitudinal data from healthcare workers

Abstract

Background:

Healthcare workers (HCW) are at high risk to develop mental health problems during the COVID-19 pandemic because of additional work load, perceived stress, and exposure to patients with COVID-19. Currently, there are few studies on change over time in the prevalence of depressive symptoms during pandemic start among HCW. Thus, the aims of the current study were to examine whether depressive symptoms increased during the pandemic and were associated with perceived stress and own COVID-19 infection and workplace exposure to virus-infected patients.

Methods:

The cohort study used longitudinal data from HCW collected monthly (July 2020 till December 2020) during the first year of the pandemic before vaccination became available. The sample of n = 166 was drawn from a German hospital and included medical (e.g. nurses, therapists, and physicians) and administrative staff. Using multilevel models, we analyzed the change in depressive symptoms [assessed with General Depression Scale (GDS), a validated German version of the Center for Epidemiological Studies Depression Scale (CES-D)] and its association with perceived stress across the study period. Laboratory-confirmed own infection was tested as a potential moderator in this context. Subscales of the GDS were used to examine change over time of depressive symptom modalities (e.g. emotional, somatic, and social interactions (β, 95% confidence interval).

Results:

Depression scores increased significantly during the study period (β = .03, 95% CI [0.02, 0.05]). Perceived stress was associated with depressive symptoms (β = .12, 95% CI [0.10, 0.14]) but did not change over time. Exposure to COVID-19 infection was associated with a higher increase of depressive symptoms (β = .12, 95% CI [0.10, .14]). Somatic symptoms of depression increased among medical HCW with workplace exposure to COVID-19 (β = .25, 95% CI [0.13, 0.38]), but not in administrators (β = .03, 95% CI [−0.04, 0.11]).

Conclusion:

Research is needed to identify factors that promote the reduction of depressive symptoms in medical HCW with exposition to COVID-19 patients. Awareness of infection protection measures should be increased.

Keywords

Depressive symptoms COVID-19 pandemic healthcare workers

Introduction

Coronavirus 2019 disease (COVID-19) has spread internationally over the past 2 years.

Since 2020, there have been waves of COVID-19 (i.e. increase of new COVID-19 cases followed by decline), the causative agent of which is ‘severe acute respiratory syndrome coronavirus 2’ (SARS-CoV-2). By this account, people faced new virus variant strains (e.g. delta variant and omicron variant) with virus-specific severity and symptoms (El-Shabasy et al., 2022). The first two peaks of coronavirus incidence had high rates of virus-related mortality (El-Shabasy et al., 2022), indicating more severe disease courses compared with the third wave. The increase in confirmed COVID-19 cases in the general population resulted in an increasing number of individuals hospitalized for COVID-19 (e.g. United States: April 20th 2020: 50 per 100,000; October 24th 2020: 202 per 100,000; COVID-Net, 2020-2021). Public health responses to contain the spread of SARS-CoV-2 included self-isolation, quarantine, school closures, curfews, and full lockdowns (Kunzler et al., 2021), which may be associated with poor mental health in individuals, and in the general public (Galea et al., 2020; Kunzler et al., 2021; Moreno et al., 2020).

Mental health can be defined as ‘a dynamic state of internal equilibrium which enables individuals to use their abilities [. . .] to cope with adverse life events and function in social roles’ (Galderisi et al., 2015). Health care workers (HCW) may be particularly vulnerable to poor mental health during the pandemic due to their working conditions and work-related distress associated with the pandemic. To date, several studies have been published on the mental health of HCW during the COVID-19 pandemic (de Kock et al., 2021; de Sousa, Tavares, et al., 2021; Marvaldi et al., 2021; Muller et al., 2020; Spoorthy et al., 2020). Previous studies operationalized mental health among HCW along with COVID-19 predominantly by outcomes related to sleep disturbance, anxiety, and depression. For example, Marvaldi et al. (2021) indicated a burden of depression and depressive symptoms, with a pooled prevalence of 31 % among 68,030 HCW. de Kock et al. (2021) summarized the findings of studies with a prevalence of depressive symptoms ranging from 9 to 50 %.

To examine protective and risk determinants that may contribute to depressive symptoms among HCW, socio-ecological predictors are discussed in the literature (Hennein et al., 2021). Briefly, socio-ecological theories attempt to explain health in the context of multiple influences at different levels, each with resources and burdens for health (Bronfenbrenner, 1979). In case of the COVID-19 pandemic, previous studies suggest that determinants at the individual and the institutional levels may contribute to depressive symptoms among HCW. For example, at an individual level young age, female sex and an own COVID-19 infection are associated with depressive symptoms (de Kock et al., 2021; Magnúsdóttir et al., 2022; Mazza et al., 2020). de Kock et al. (2021) found that medical HCW (e.g. physicians and nurses) have a higher burden of depressive symptoms than nonmedical HCW (institutional level). In addition, HCW who have job-related exposure to virus-infected patients are thought to be at risk for mental health burdens, due to their workload, inadequate personal protective equipment, or decisions to triage COVID-19 patients (de Kock et al., 2021; Giang et al., 2020; Lu et al., 2020; Moreno et al., 2020; Nicolaou et al., 2021; Zhang et al., 2020). As such, institutional organization may contribute to concerns about own infection (Cai et al., 2020) and psychological distress at an individual level (Muller et al., 2020) that, in turn, contribute to depressive symptoms (de Sousa, Vargas, et al., 2021).

Mental health impairment in HCW is widely recognized as an important public health and health care challenge in the pandemic, as it is associated with both immediate and long-term consequences in individuals and for health care. For example, poor mental health contribute to sick absence from work (Spoorthy et al., 2020), medical errors and turnover rates (Marvaldi et al., 2021), as well as to severity of an acute own COVID-19 infection (Magnúsdóttir et al., 2022; Xie et al., 2022) with elevated levels of mental sequelae (Zeng et al., 2022). Promoting health of HCW is considered critical in pandemic management to health and to continuity of healthcare provision (Adams & Walls, 2020; Sahu et al., 2020; World Health Organisation [WHO], 2020, 2022).

The current literature is insufficient to comprehensively assess the impact of the pandemic on depressive symptoms in HCW and associated determinants for several reasons. First, recent findings on the prevalence of depressive symptoms in HCW are contradictory (de Kock et al., 2021; Marvaldi et al., 2021), which may depend on the operationalization of the outcome, the sample characteristics, and/or measurement instruments. Second, existing studies on depressive symptoms at the beginning of the pandemic among HCW contain few longitudinal data of outcomes or determinants (Jordan et al., 2023). Third, determinants or moderating factors of depressive symptoms over time that account for actual workplace exposure to COVID-19 were not examined.

The aim of the current study was to evaluate a change in the prevalence of depressive symptoms in HCW during the pandemic. According to the ecological theory mentioned above, perceived stress and healthcare workers’ own infection status (individual level), as well as workplace exposure to virus-infected patients (institutional level) were considered to explore the following research questions:

RQ1: Was there a change in depressive symptoms among HCW during the first wave of the COVID-19 pandemic, and was this change associated with perceived stress?

RQ2: Is change in depressive symptoms moderated by HCW’s own COVID-19 infection and/or workplace exposure to COVID-19 patients?

RQ3: Do depressive symptom modalities differ by workplace exposure to COVID-19 patients?

Methods

Data from a prospective cohort study

We used data from a prospective cohort study conducted in a standard care hospital at the primary healthcare level in the German federal state of Brandenburg. The aim of this study was to examine (i) the seroprevalence of antibodies against SARS-CoV-2 among HCW and (ii) self-reported exposure to COVID-19 patients at the job, as well as (iii) subjective physical and mental health. For this purpose, we recruited n = 166 participants from different hospital wards (i.e. emergency department, intensive care unit, cardiology, geriatrics, pediatrics, laboratory, radiology, and administration). To avoid possible selection bias, all persons currently working in the hospital wards concerned were to be recruited. The wards were selected according to suspected exposure to COVID-19 patients. This strategy made it possible to include the majority of eligible staff in the sample, who were predominantly female and on average 45 years old. This is consistent with German hospitals 2021 (about 80% female; age <34 years (35.2%), >35 to 44 years (20.8%), >45 to 54 years (21.7%), >55 years (22.4%; Fuchs & Weyh, 2023; Wasem & Blase, 2023)). Blood samples for laboratory analyses and questionnaires were collected monthly from July 2020 to December 2020. Thus, the data reflected the first two peaks of coronavirus incidence in the basic population in Brandenburg, when licensed vaccines were not yet available and the ‘sense of vulnerability’ was increasing among people, in part due to rising mortality rates (El-Shabasy et al., 2022).

Further study information (e.g. participation rates) and results of serological analysis were previously published (Hoffmann et al., 2021). Our study was ethically approved according to German ethical guidelines and informed consent was obtained by participants.

Measures

Main dependent variable: Repeatedly measured mean continuous depression scores (GDS mean scores)

Our main outcome of interest was self-reported mental health, operationalized by depressive symptoms. We administered the validated General Depression Scale (GDS, reliability (Cronbach’s alpha) = .89–.92, convergent validity e.g. with Becks inventory for measuring depression = .88 (Beck et al., 1961; Hautzinger et al., 2012; Shafer, 2006), which is the German version of the ‘Center for Epidemiological Studies Depression Scale’ (Carleton et al., 2013; Radloff, 1977). The GDS comprises 20 items of self-descriptive statements of oneself, each in relation to the last week and each on a 4-point scale (0: less than 1 day, 1: 1–2 days, 2: 3–4 days, and 3: 5–7 days). GDS mean scores were calculated for all participants with three or fewer missing items at each survey timepoint, in line with previous studies (Kleih et al., 2022). Higher mean scores indicated more depressive symptoms, with a range of values of 0.00 to 2.30 in this study (GDS range of values: 0.00–3.00).

Second dependent variable: Repeatedly measured continuous scores of GDS subscales

Depressive symptoms manifest in different modalities (e.g. emotional, somatic, and interactional). In order to evaluate temporal effects, we computed five subscales based on all GDS items (n = 20) that have been previously confirmed in factor analytic studies (Hautzinger et al., 2012):

(1) Emotional symptoms: item numbers 1, 3, 6, 10, 12, 14, and 18 (n = 7)

(2) Motivation symptoms: item numbers 7 and 20 (n = 2)

(3) Cognitive symptoms: item numbers 4, 5, 8, 9, and 16 (n = 5)

(4) Somatic symptoms: item numbers 2, 11, and 17 (n = 3)

(5) Interactional symptoms: item numbers 13, 15, and 19 (n = 3)

Predictor variables (independent variables)

a. Linear time

The variable time specified whether the depression scores were collected at baseline (July 2020), ‘Time 1’ (August 2020), ‘Time 2’ (September 2020), ‘Time 3’ (October 2020), ‘Time 4’ (November 2020), or at ‘Time 5’ (December 2020). Since we collected data at more than three timepoints, we tested time as a linear predictor variable (Twisk, 2006).

b. Previous own infection among HCW

A previous own infection with SARS-CoV-2 among HCW was confirmed by laboratory analysis with serological testing of total Ig antibodies against SARS-CoV-2. Briefly, a nucleocapsid antigen from a SARS-CoV-2 assay was applied, resulting in a cut-off index on a continuous scale (99.5% sensitivity and 99.8% specificity). Further information on applied laboratory analytics were previously published (Hoffmann et al., 2021).

We converted the values to a dichotomous variable, with a cut-off index (COI) ⩾1.00 interpreted as positive for SARS-CoV-2 antibodies (i.e. confirmed own infection).

c. Perceived stress among HCW

We assessed self-rated stress among HCW using the 4-item version of the validated Perceived Stress Scale (Cohen et al., 1983; PSS). This involves a critical appraisal of one’s life in the last month according to the extent of and control over stressful situations, with each item rated on a 5-point scale (0 – never, 1 – almost never, 2 – sometimes, 3 – fairly often, and 4 – very often). The scores of the individual items were summed, yielding a range of values from 0 to 16, with higher scores indicating a higher amount of perceived stress.

We tested a correlation of PSS4 sum scores with symptoms of depression for each survey timepoint. Furthermore, we examined a linear time trend in PSS4 sum scores, and they neither increase nor decrease (i.e. remain stable) over time among participants. The Supplemental Material presents corresponding data (Supplemental Tables 4 and 5). Therefore, for analysis, mean stress scores were computed using PSS4 for participants for whom the complete scale items were collected at a minimum of 3 timepoints. Accordingly, this variable (stress) reflects an average of PSS scores across the entire study period.

Covariates

Sociodemographic data, including information on sex [female/male] and age [year of birth], was collected using a questionnaire that participants completed at baseline.

We clustered participants into four groups according to which hospital ward they were working at: 1 – emergency department; intensive care unit; 2 – cardiology, geriatrics, pediatrics; 3 – laboratory, radiology; and 4 – administration (Hoffmann et al., 2021). We then built a dichotomous variable ‘group’ [administrative/medical staff] indicating exposure to workplace contact with COVID-19 patients, presuming that medical staff have more frequent contact than administrative staff.

Statistical analyses

All analyses were performed using R (version 4.0.3 2022). All tests were considered significant at p < .05 and were stratified by ‘group’ (i.e. indicating varying degrees of workplace exposure to COVID-19 patients).

In order to perform a ‘complete case analysis’ (Lewin et al., 2018), we created an analytic dataset with missing values omitted for each variable used (i.e. scores of depressive symptoms at all timepoints, age, sex, COI value, and average PSS scores). Differences between medical and administrative staff in workplace exposure to COVID-19 patients were tested using Fisher’s exact test (Kim, 2017).

Step 1: Confirmation of a linear time trend in depressive symptoms and a predictive value of perceived stress

In the first step, we attempted to confirm a linear time trend in the change in depressive symptoms. We ran three (M0–M2) linear multilevel models with repeatedly measured mean depression scores as the dependent variable by using lme() function in R (Laird & Ware, 1982; Pinheiro et al., 2022). In these models, we included time as a linear predictor by using it as a fixed effect, which specifies whether mean depression scores were measured at baseline (July 2020), ‘Time 1’ (August 2020), ‘Time 2’ (September 2020), ‘Time 3’ (October 2020), ‘Time 4’ (November 2020), or at ‘Time 5’ (December 2020). To account for differences in the dependent variable both within level (level 1: data at different timepoints) and between levels (level 2: individuals), we used ‘subject’ as a random effect. The variable ‘subject’ indicated whether data belong to the same person. Since we analyzed repeatedly measured data, we accounted for residual autocorrelation of the data between survey timepoints by extending the models through the introduction of a covariance structure. We used the predefined first-order autoregressive structure (AR1) as recommended by Horváth et al. (2014) for linear time series data analysis.

We used the following equation to predict repeatedly measured scores of depressive symptoms by time and stress, while adjusting for age and sex:

\begin{array}{l} Mean depression {score}_{t, n} = β i * {time}_{t, n} + β n * {stress}_{t, n} + β n * {age}_{t, n} + β n * {sex}_{t, n} \\ t time (Level 1) n subject (Level 2) \end{array}

Using ANOVA, we investigated differences in the model fit by comparison of the Akaike information criterion (AIC value; Wagenmakers & Farrell, 2004)):

- Model M0: random intercept (level 2: ‘subject’)

- Model M1: random intercept, with time’ and stress as predictors (level 1: time, stress, level 2: ‘subject’)

- Model M2: random intercept with first-order autoregressive covariance structure (AR1), with time and stress as predictors (level 1: time, stress, level 2: ‘subject’)

Testing for multicollinearity (Lüdecke et al., 2021): We found no evidence for significant multicollinearity with variance inflation factors (VIF) below 1.07 for all predictors and covariates (time, age, sex, stress). Furthermore, for all variables the average VIF was <1.1 and tolerance statistics were >.93 (Bowerman, 2000).

Testing for heterogeneity biased estimates and inferential statistics (Bell & Jones, 2015): Using the function ‘check_heterogeneity_bias’ (performance package; (Lüdecke et al., 2021)), we found no evidence for significant correlation of age, sex, and stress, each with the random effect (‘subject’).

Step 2: Differences in the linear time trend in depressive symptoms by own infection and/or by workplace exposure to COVID-19 patients?

As described above, theories about the pathways through which mental health may be affected suggest that a previously confirmed viral infection may contribute to poorer outcomes (Mazza et al., 2020). Thus, we evaluated whether a previously confirmed own virus infection predicts mean depression scores over time. We extended Model M2 with an interaction term time × infection for Model M3, where infection indicates whether a SARS-CoV-2 infection could be newly confirmed at a survey timepoint among HCW.

We repeated Model M2 and Model M3 stratified by workplace exposure to COVID-19 patients. Significant interactions are graphically depicted (see Figure 1).

Figure 1.

Predicted values of depressive symptoms by laboratory-confirmed infection status among overall sample (infection 0 = no, 1 = yes).

Step 3: Differences in the linear time trend in depressive modalities by presumed workplace exposure to COVID-19 patients

Recent literature suggests mental health burden (e.g. anxiety, depression, sleep problems, and distress; Muller et al., 2020) varies by workplace exposure to COVID-19 patients (Muller et al., 2020). To evaluate a temporal change in modalities, we reran the above Model M2, with fitting each of the GDS subscales as dependent variables to the linear multilevel models.

Results

Of n = 166 healthcare workers recruited, a total of n = 91 had complete data on mean depression scores and own infection status at each survey timepoint as well as baseline information (i.e. age, sex, and hospital ward group). In addition, these HCW had complete PSS data at a minimum of three survey timepoints.

Sample characteristics of these n = 91 participants are displayed in Table 1. Administrative staff mainly reflect individuals with administrative and laboratory jobs (about 72%), while those for medical staff refer to nurses, physicians, and physical therapists (about 90%). While among medical HCW, 74.47% (n = 35) reported workplace exposure to COVID-19 patients at time 5, this proportion was 29.55% (n = 13) among administrative staff. The more frequent exposure among medical HCW was statistically significant (p = .014).

Table 1.

Sample characteristics (n = 91).

	Group: administrative staff	Group: medical staff
% (N)	48.35 (44)	51.65 (47)
Age, mean in years (SD)	46.52 (8.4)	44.38 (12.21)
Female sex % (N)	88.64 (39)	91.49 (43)
Profession	% (N)	% (N)
Administrative staff	52.27 (23)	4.26 (2)
Social service staff	4.55 (2)	0.00 (0)
Radiology assistant	9.09 (4)	2.13 (1)
Lab assistant	20.45 (9)	0.00 (0)
Technical assistant	2.27 (1)	0.00 (0)
Patient service staff	2.27 (1)	4.26 (2)
Physical therapist	2.27 (1)	12.77 (6)
Nursing staff	6.82 (3)	63.83 (30)
Physician	0 (0)	12.77 (6)
Depressive symptoms score	Mean (SD)	Mean (SD)
‘Time 0’	0.48 (0.42)	0.56 (0.41)
‘Time 1’	0.49 (0.38)	0.45 (0.37)
‘Time 2’	0.47 (0.42)	0.44 (0.42)
‘Time 3’	0.45 (0.39)	0.46 (0.39)
‘Time 4’	0.50 (0.48)	0.65 (0.48)
‘Time 5’	0.62 (0.59)	0.77 (0.37)
Perceived stress score	Mean (SD)	Mean (SD)
‘Time 0–5’	4.92 (2.67)	5.38 (2.43)
Laboratory confirmed own infection	% (N)	% (N)
‘Time 0–5’	9.09 (4)	25.53 (12)
Workplace exposure to COVID-19 patients	% (N)	% (N) p-value test statistic
‘Time 0’	4.55 (2)	12.77 (6)
‘Time 1’	0.00 (0)	6.38 (3)
‘Time 2’	0.00 (0)	2.13 (1)
‘Time 3’	11.36 (5)	25.53 (12) p = .051
‘Time 4’	25.00 (11)	46.81 (22) p = .029
‘Time 5’	29.55 (13)	74.47 (35) p = .014

Step 1: Confirmation of a predictive value of both time and stress on depressive symptoms

Model M0 in Table 2 shows 62% explained variance (ICC) by ‘subject’, indicating significant between-level variation requiring multilevel analyses.

Table 2.

Linear time and stress as predictors of mean depression scores among healthcare workers.

	Group administrative staff	Group medical staff	Overall sample
	Subject n = 44	Subject n = 47	Subject n = 91
	Obs. n = 264	Obs. n = 282	Obs. n = 564
Model 0 (random intercept):
subject:	ICC = .67	ICC = .58	ICC = .62
	AIC = 105.29	AIC = 237.82	AIC = 352.03
Model 1 (Model 0 + linear time):
time:	β = .02*(.00–.04)	β = .05***(.03–.07)	β = .03***(.02–.05)
stress:	β = .12***(.10–.14)	β = .13***(.10–.16)	β = .12***(.10–.14)
marginal R²:	.573	.459	.493
	AIC = 41.40^a	AIC = 174.72^a	AIC = 228.55^a
Model 2 (Model 1 + first-order autoregressive covariance structure)
time:	β = .02*(.00–.04)	β = .04***(.02–.07)	β = .03***(.02–.05)
stress:	β = .12***(.10–.14)	β = .14***(.11–.17)	β = .12***(.10–.14)
marginal R²:	.574	.460	.495
	AIC = 29.94^a,b	AIC = 154.94^a,b	AIC = 191.75^a,b
Model 3 (Model 2 + time × own infection):
time:	β = .02 (−.00–.04)	β = .03*(.00–.06)	β = .02**(.01–.04)
time × infection:	β = .07^~ (−.00–.14)	β = .05 (−.01–.10)	β = .06*(.01–.10)
stress:	β = .12***(.10–.14)	β = .13***(.10–.16)	β = .12***(.10–.14)
marginal R²:	.586	.470	.502
	AIC = 29.12^a,b	AIC = 155.54^a,b	AIC = 187.13^a,b,c

Note. All analyses are adjusted by sex and age; β = unstandardized estimate with 95% confidence intervals in brackets; AIC = akaike information criterion; ICC = intra class correlation; Obs. = Observations.

Significantly better fit than model 0 p < .001.

Significantly better fit than model 1 p < .001.

Significantly better fit than model 2 p < .05.

p < .10. *p < .05. **p < .01. ***p < .001.

Time was significantly associated with mean depression scores in Model M1. Similarly, model fit improved significantly compared to Model M0. Adding a first-order autoregressive covariance structure to Model M1 resulted in a further significant improvement of Model M2. The prediction of mean depression scores by time (β = .03) was significant (95% CI [0.02, 0.05]), indicating an increase of depressive symptoms over time in the overall sample. In Model M2, stress significantly predicted repeatedly measured depressive symptoms (β = .12, 95% CI [0.10, 0.14]).

Step 2: Differences in the linear time trend in repeatedly measured depressive symptoms by own COVID-19 infection among HCW and/or by workplace exposure to COVID-19 patients

Model M3 showed a significant interaction with a laboratory-confirmed own infection among overall sample and, likewise, had a better model fit than Model M0, Model M1, and Model M2 (see Table 2).

Figure 1 depicts the significant interaction between a linear time predictor in mean depression scores and an infection status among overall sample. Among individuals with confirmed infection status, there was a stronger positive relationship between time and depressive symptom scores.

When stratified by workplace exposure to COVID-19 patients, Model M2 maintained a predictive value of time for depressive symptoms among medical HCW (β = .04, 95% CI [0.02, 0.07]), and a marginal effect among administration staff (β = .02, 95% CI [0.00, 0.04]). Similarly, model fit of Model M2 improved compared to Model M0 and Model M1 for both groups. Additionally, Model M2 retained a predictive value of stress on depressive symptoms among medical HCW (β = .14, 95% CI [0.11, 0.17) and administration staff (β = .12, 95% CI [0.10, 0.14]). Model M3 did not maintain an interaction effect among health care worker groups. Accordingly, Model M3 did not show improved model fit in the analysis of stratified sample data.

Step 3: Differences in the linear time trend in repeatedly measured modalities by workplace exposure to COVID-19 patients

Of the analysis sample (n = 91), 71 participants answered all items in each GDS subscale at each survey timepoint, which was considered valid for computing the subscale scores.

Table 3 shows a replication of Model M2 with repeatedly measured sums in the GDS subscales as dependent variables. A predictive value of time maintained for emotional symptoms among both administrative (β = .31, 95% CI [0.14, 0.49]) and medical HCW (β = .42, 95% CI [0.14, 0.70]). Somatic symptoms increased significantly over time among medical HCW (β = .25, 95% CI [0.13, 0.38]), but not among administrative staff (β = .03, 95% CI [−0.04, 0.11]). Model M2 retained a marginal effect of time for motivational (β = .06, 95% CI [0.00, 0.13]) and interactional symptoms (β = .10, 95% CI [0.00, 0.12]) among administrative staff, but not among medical HCW.

Table 3.

Linear time and stress as predictors of depressive modalities among healthcare workers by workplace exposure to COVID-19 patients.

	Group administrative staff	Group medical staff
	Subject n = 37	Subject n = 34
	Obs. n = 222	Obs. n = 204
AR1 – Model 4: emotional symptoms
time	β = .31***(.14–.49)	β = .42**(.14–.70)
stress	β = .92***(.70–1.13)	β = .91***(.64–1.19)
marginal R²	.490	.363
AR1 – Model 5: motivational symptoms
time	β = .06^~(.00–.13)	β = .05 (−.03–.14)
stress	β = .32***(.24–.41)	β = .32***(.23–.41)
marginal R²	.420	.346
AR1 – Model 6: cognitive symptoms
time	β = .11 (−.02–.24)	β = .06 (−.12–.24)
stress	β = .72***(.50–.94)	β = .90***(.68–1.12)
marginal R²	.461	.457
AR1 – Model 7: somatic symptoms
time	β = .03 (−.04–.11)	β = .25***(.13–.38)
stress	β = .33***(.23–.43)	β = .35***(.23–.46)
marginal R²	.352	.302
AR1 – Model 8: interactional symptoms
time	β = .10^~(.00–.12)	β = .02 (−.04–.10)
stress	β = .26***(.15–.38)	β = .22**(.08–.35)
marginal R²	.177	.134

Note. All analyses are adjusted by gender and age; β = unstandardized estimate with 95% confidence intervals in brackets; AR1 = first-order autoregressive covariance structure; Obs. = observations.

p < 10. **p < .01. ***p < .001.

Perceived stress predicted the scores for each repeatedly measured depressive symptom modality and in each HCW group.

Discussion

The present study examined longitudinal changes in depressive symptoms and the predictive value of perceived stress among healthcare workers in 2020, taking into account both workplace exposure to COVID-19 patients and laboratory-confirmed infection status. Analyses were based on data collected at a general primary care hospital during the first two peaks of coronavirus incidence in the population in the German federal state of Brandenburg, when no licensed vaccine was available yet.

RQ1: Temporal increase in mean values of depressive symptoms and predictive value of perceived stress during the first wave of the COVID-19 pandemic

Multilevel models showed a temporal increase in depressive symptoms among healthcare workers. These results are in line with previous findings of the population-based ‘Gesundheit in Deutschland aktuell’ study (GEDA for short; engl. ‘Health in Germany up-to-date study’) conducted in 2014/2015 by the Robert Koch-Institute on the prevalence of depressive symptoms in the general population of 10.1 %, even beyond clinical diagnoses (Bretschneider et al., 2017). The temporal trend in the increase of depressive symptoms is partially consistent with studies showing an increase in the prevalence of self-reported clinical diagnoses of depression according to the past 12 months in recent years (Thom et al., 2017). However, these findings are comparable to a limited extent as Thom et al. (2017) did refer to a different operationalization of depressiveness (‘ever medically diagnosed’ vs. here: GDS, continuous measure of depressive symptomatology). By contrast, Damerow et al. (2022) confirmed a temporal decrease of depressive symptoms in the period from April to August 2020 in the general German population based on GEDA data, which have been collected in cross-section in 2019/2020. The authors argue that the COVID-19 pandemic did not contribute to an increase in depressive symptoms in 2020 (Damerow et al., 2022). However, comparability of results is limited due to both the study population (here: HCW) and the data (here: longitudinal).

Multilevel models showed an additional predictive value of perceived stress on changes in depressive symptoms, an association that confirms findings from previous studies on stress and depression in HCW during the pandemic (Li et al., 2022) and that has been discussed in a recent narrative review (de Sousa, Vargas, et al., 2021).

RQ 2: Own infection exacerbates temporal increase in mean depressive symptoms, regardless of workplace exposure to COVID-19 patients

When stratified by workplace exposure to COVID-19 patients, time and stress additively predicted mean depression scores, but without significant differences between HCW groups. By analyzing the moderating effect of a previous infection, we found an additional interaction effect between time and infection in the overall sample, but not when stratified by workplace exposure to COVID-19 patients, suggesting a steaper increase in depression symptom scores over time among individuals who were infected with SARS-CoV-2 and those who were not infected. The recent study confirms previous findings on mental health burden in individuals with COVID-19. However, previous studies on the association between mental health and COVID-19 were mainly descriptive in nature or reflect retrospective case control or cohort study designs (de Kock et al., 2021; WHO, 2022). Briefly, poor mental health in individuals may both predispose them to severe COVID-19 disease (WHO, 2022) and be associated with mental health symptoms as a consequence of a COVID-19 diagnosis (Magnúsdóttir et al., 2022; Xie et al., 2022). Our findings based on longitudinal data extend previous findings by suggesting a directed association between infection and poor mental health (here: depressive symptoms). Hence, our findings support previous results of cross-sectional data analyses indicating COVID-19 to be associated with the onset of mental disorders (e.g. anxiety, depression, and stress) after an acute illness (Magnúsdóttir et al., 2022; Xie et al., 2022).

Our results support the following suggestions about mechanisms by which COVID-19 infections influence depression symptoms. First, previous studies pointed to psychological effects resulting from, among other things, social isolation due to one’s infection. For example, Pancani et al. (2021) found physical separation to be associated with depression. Second, studies pointed to inflammatory processes as a biological mechanism underlying the onset of depression (Mazza et al., 2020; Miller & Raison, 2016). In line with this, we found that scores for depressive symptoms were higher over time when an infection was previously confirmed. Consequently, risk for depression in the post-disease phase could be affected by infection-related demands on the immune system (Mazza et al., 2020; Miller & Raison, 2016).

RQ 3: Workplace exposure to COVID-19 patients affects temporal increase in somatic depressive symptoms

Multilevel models showed an increase in depressive symptom modalities according to workplace exposition to COVID-19 patients. In both HCW groups, emotional symptoms (i.e. feelings of worry, gloom, dejection, anxiety, loneliness, and sadness) increased over time. Studies of mental health determinants among HCW at the beginning of the pandemic suggest that they may be concerned about infecting themselves or their families with COVID-19 or about rising mortality rates (Cai et al., 2020; El-Shabasy et al., 2022) or they may increasingly experienced infections by their own or infections and deaths in relatives and acquaintances.

In particular, medical staff with increasing patient contact showed increasing scores for depression-related somatic symptoms (i.e. fatigue, sleep disturbance, and weepiness), whereas this was not the case for administrative staff. These findings are in line with evidence demonstrating associations between hospital working ward and mental health among HCW. For instance, Lai et al. (2020) found that HCW involved in the frontline diagnosis, treatment, and care of patients with COVID-19 are at higher risk of developing depression compared to administration staff. Sagherian et al. (2020) indicated that hospital nursing staff suffered from poor sleep and fatigue during the COVID-19 pandemic. Our findings revealed an additively predictive value of perceived stress on somatic and emotional symptoms. This finding points to individual and institutional determinants of depressive symptoms and, thus, supports the above ecological theory. For example, Sagherian et al. (2020) refered to an institutional level by pointing to decreased intershift recovery or avoiding breaks during healthcare provision in the first wave of the pandemic that, in turn, are associated with quality of sleep (Xiao et al., 2020) and perceived stress on an individual level (Spoorthy et al., 2020; Xiao et al., 2020).

Strength and limitations

To the best of our knowledge, this study is one of the first to use longitudinal data analysis to examine the change in depressive symptoms recorded repeatedly over a 6-month period at the beginning of the pandemic. Our main outcome was assessed using the GDS, a validated measure of depressive symptoms. In addition, previous infection with SARS-CoV-2 was confirmed by laboratory analysis and included in our analysis. A further strength of the study is that it was conducted during the early phase of the pandemic when vaccines were not yet available, thus findings are not confounded by vaccination status.

Our study has several limitations. First, due to item nonresponse across all survey timepoints, we were only able to include n = 564 observations from 91 participants. However, this subsample was similar in age and sex compared the original sample. Eight participants with an infection with SARS-CoV-2 dropped from the recruited sample due to missing data (Hoffmann et al., 2021). The current analysis is based on data collected at six time points in 2020 from HCW at a primary care hospital. Further research could consider longitudinal research on mental health among HCW in different healthcare settings (e.g. university hospitals, rehabilitation hospitals, or outpatient care) or at different points during the pandemic (de Kock et al., 2021). Finally, we were unable to include additional potential sources for psychological stress (e.g. social isolation; Brooks et al., 2020) due to missing responses in the questionnaires or a lack of variance in the data. Since recent findings suggested that socio-ecological factors are helpful to examine health among HCW (Hennein et al., 2021), such factors (e.g. team cohesion and hospital politics) should be considered in future longitudinal studies of HCW’s mental health.

Conclusion

In the present study, the prevalence of depressive symptoms in healthcare workers was found to increase between the first two incidences peaks of COVID-19 cases in 2020. Our study revealed that newly confirmed own infection contributed to the temporal increase of depressive symptoms among healthcare workers during the study period, regardless of the workplace exposition to COVID-19 patients. In addition, depression-related somatic symptoms increased over time among HCW that provided direct care to COVID-19 patients, but not among administrative staff. We conclude that job-related contact with COVID-19 and an own infection predict deterioration of mental health and related symptoms among HCW over time. Our findings could help to improve the mental health of HCW by emphasizing both the need for personal protective measures against infectious diseases and workplace health promotion in clinical settings. Further research is needed to identify protective factors that promote the reduction of depressive symptoms in medical HCW with exposition to COVID-19 patients, including stress-related determinants.

Supplemental Material

sj-docx-1-isp-10.1177_00207640231196737 – Supplemental material for Did the prevalence of depressive symptoms change during the COVID-19 pandemic? A multilevel analysis on longitudinal data from healthcare workers

Supplemental material, sj-docx-1-isp-10.1177_00207640231196737 for Did the prevalence of depressive symptoms change during the COVID-19 pandemic? A multilevel analysis on longitudinal data from healthcare workers by Stephanie Hoffmann, Susanne Schulze, Antje Löffler, Juliane Becker, Frank Hufert, Heinz-Detlef Gremmels, Christine Holmberg, Michael A Rapp, Sonja Entringer and Jacob Spallek in International Journal of Social Psychiatry

Footnotes

Author contributions

Conceptualization: JS, SH, JB, FH, HDG, and SE; Data collection: JB and SH; Methodology: SH, JS, and SE; Data curation: SH and SS; Formal analysis: SH; Writing – original draft: SH; Writing – review and editing: SH, SE, SS, JB, JS, HDG, AL, CH, MR, and FH; Supervision: JS.

Conflict of interest

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

The ethics committee of the Brandenburg University of Technology Cottbus-Senftenberg (EK2020-8) gave ethical approval for complete study. Written, informed consent for participation in this study was obtained from all participants.

ORCID iDs

Stephanie Hoffmann

Susanne Schulze

Christine Holmberg

Jacob Spallek

Supplemental material

Supplemental material for this article is available online.

References

Adams

J. G.

Walls

R. M.

(2020). Supporting the health care workforce during the COVID-19 global epidemic. JAMA, 323, 1439–1440. https://doi.org/10.1001/jama.2020.3972

Beck

Ward

Mendelson

Mock

Erbaugh

(1961). An inventory for measuring depression. Archives of General Psychiatry, 4, 561–571. https://doi.org/10.1001/archpsyc.1961.01710120031004

Bell

Jones

(2015). Explaining fixed effects: Random effects modeling of time-series cross-sectional and panel data. Political Science Research and Methods, 3, 133–153. https://doi.org/10.1017/psrm.2014.7

Bowerman

B. L.

(2000). Linear statistical models: An applied approach. Thomson Learning.

Bretschneider

Kuhnert

Hapke

(2017). Depressive Symptomatik bei Erwachsenen in Deutschland [Depressive symptoms among adults in Germany]. Journal of Health Monitoring, 2, 77–83.

Bronfenbrenner

(1979). The ecology of human development: Experiments by nature and design (p. 330). Harvard University Press.

Brooks

S. K.

Webster

R. K.

Smith

L. E.

Woodland

Wessely

Greenberg

Rubin

G. J.

(2020). The psychological impact of quarantine and how to reduce it: Rapid review of the evidence. The Lancet, 395, 912–920. https://doi.org/10.1016/S0140-6736(20)30460-8

Cai

Chen

Jiang

Zhuang

(2020). Psychological impact and coping strategies of frontline medical staff in Hunan between January and March 2020 during the outbreak of coronavirus disease 2019 (COVID‑19) in Hubei, China. Medical Science Monitor: International Medical Journal of Experimental and Clinical Research, 26, e924171. https://doi.org/10.12659/MSM.924171

Carleton

R. N.

Thibodeau

M. A.

Teale

M. J. N.

Welch

P. G.

Abrams

M. P.

Robinson

Asmundson

G. J. G.

(2013). The Center for Epidemiologic Studies Depression Scale: A review with a theoretical and empirical examination of item content and factor structure. PLoS One, 8, e58067. https://doi.org/10.1371/journal.pone.0058067

10.

Cohen

Kamarck

Mermelstein

(1983). A global measure of perceived stress. Journal of Health and Social Behavior, 24, 385. https://doi.org/10.2307/2136404

11.

COVID-Net. (2020-2021). Laboratory-confirmed COVID-19-associated hospitalizations. Retrieved June 8, 2022, from https://gis.cdc.gov/grasp/COVIDNet/COVID19_3.html

12.

Damerow

Rommel

Beyer

A.-K.

Hapke

Schienkiewitz

Starker

Richter

Baumert

Fuchs

Gaertner

Müters

Lemcke

Allen

(2022). Gesundheitliche Lage in Deutschland in der COVID-19-Pandemie. Zeitliche Entwicklung ausgewählter Indikatoren der Studie GEDA 2019/2020-EHIS – Ein Update [Health situation in Germany during the COVID-19 pandemic. Development over time of selected indicators of the GEDA 2019/2020-EHIS study - An update]. Robert Koch Institute. https://doi.org/10.25646/9880

13.

de Kock

J. H.

Latham

H. A.

Leslie

S. J.

Grindle

Munoz

S.-A.

Ellis

Polson

O’Malley

C. M.

(2021). A rapid review of the impact of COVID-19 on the mental health of healthcare workers: Implications for supporting psychological well-being. BMC Public Health, 21, 104. https://doi.org/10.1186/s12889-020-10070-3

14.

de Sousa

G. M.

Tavares

V. D. O.

de Meiroz Grilo

M. L. P.

oelho

M. L. G.

de Lima-Araújo

G. L.

Schuch

F. B.

Galvão-Coelho

N. L.

(2021). Mental health in COVID-19 pandemic: A meta-review of prevalence meta-analyses. Frontiers in Psychology, 12, 703838. https://doi.org/10.3389/fpsyg.2021.703838

15.

de Sousa

G. M.

Vargas

H. D. Q.

Barbosa

F. F.

Galvão-Coelho

N. L.

(2021). Stress, memory, and implications for major depression. Behavioural Brain Research, 412, 113410. https://doi.org/10.1016/j.bbr.2021.113410

16.

El-Shabasy

R. M.

Nayel

M. A.

Taher

M. M.

Abdelmonem

Shoueir

K. R.

Kenawy

E. R.

(2022). Three waves changes, new variant strains, and vaccination effect against COVID-19 pandemic. International Journal of Biological Macromolecules, 204, 161–168. https://doi.org/10.1016/j.ijbiomac.2022.01.118

17.

Fuchs

Weyh

(2023). The labour market situation in hospitals. In Klauber

Wasem

Beivers

Mostert

(Eds.), Hospital report 2023 (1st ed., pp. 33–47). Springer Berlin/Heidelberg.

18.

Galderisi

Heinz

Kastrup

Beezhold

Sartorius

(2015). Toward a new definition of mental health. World Psychiatry, 14, 231–233. https://doi.org/10.1002/wps.20231

19.

Galea

Merchant

R. M.

Lurie

(2020). The mental health consequences of COVID-19 and physical distancing: The need for prevention and early intervention. JAMA Internal Medicine, 180, 817–818. https://doi.org/10.1001/jamainternmed.2020.1562

20.

Giang

T.-L.

D.-T.

Vuong

Q.-H.

(2020). COVID-19: A relook at healthcare systems and aged populations. Sustainability, 12, 4200. https://doi.org/10.3390/su12104200

21.

Hautzinger

Bailer

Hofmeister

Keller

(2012). ADS: General Depression Scale. Manual (2nd ed.). Hogrefe.

22.

Hennein

Mew

E. J.

Lowe

S. R.

(2021). Socio-ecological predictors of mental health outcomes among healthcare workers during the COVID-19 pandemic in the United States. PLoS One, 16, e0246602. https://doi.org/10.1371/journal.pone.0246602

23.

Hoffmann

Schiebel

Hufert

Gremmels

H.-D.

Spallek

(2021). COVID-19 among healthcare workers: A prospective serological-epidemiological cohort study in a standard care hospital in rural Germany. International Journal of Environmental Research and Public Health, 18, 10999. https://doi.org/10.3390/ijerph182010999

24.

Horváth

Kokoszka

Rice

(2014). Testing stationarity of functional time series. Journal of Econometrics, 179, 66–82. https://doi.org/10.1016/j.jeconom.2013.11.002

25.

Jordan

J.-A.

Shannon

Browne

Carroll

Maguire

Kerrigan

Hannan

McCarthy

Tully

M. A.

Mulholland

Dyer

K. F. W.

(2023). Healthcare staff mental health trajectories during the COVID-19 pandemic: Findings from the COVID-19 Staff Wellbeing Survey. BJPsych Open, 9, e112. https://doi.org/10.1192/bjo.2023.497

26.

Kim

H.-Y.

(2017). Statistical notes for clinical researchers: Chi-squared test and Fisher’s exact test. Restorative Dentistry & Endodontics, 42, 152–155. https://doi.org/10.5395/rde.2017.42.2.152

27.

Kleih

T. S.

Entringer

Scholaske

Kathmann

DePunder

Heim

C. M.

Wadhwa

P. D.

Buss

(2022). Exposure to childhood maltreatment and systemic inflammation across pregnancy: The moderating role of depressive symptomatology. Brain, Behavior, and Immunity, 101, 397–409. https://doi.org/10.1016/j.bbi.2022.02.004

28.

Kunzler

A. M.

Röthke

Günthner

Stoffers-Winterling

Tüscher

Coenen

Rehfuess

Schwarzer

Binder

Schmucker

Meerpohl

J. J.

Lieb

(2021). Mental burden and its risk and protective factors during the early phase of the SARS-CoV-2 pandemic: Systematic review and meta-analyses. Globalization and Health, 17, 34. https://doi.org/10.1186/s12992-021-00670-y

29.

Lai

Wang

Cai

Wei

Chen

Tan

Kang

Yao

Huang

Wang

Liu

(2020). Factors associated with mental health outcomes among health care workers exposed to coronavirus disease 2019. JAMA Network Open, 3, e203976. https://doi.org/10.1001/jamanetworkopen.2020.3976

30.

Laird

N. M.

Ware

J. H.

(1982). Random-effects models for longitudinal data. Biometrics, 38, 963. https://doi.org/10.2307/2529876

31.

Lewin

Brondeel

Benmarhnia

Thomas

Chaix

(2018). Attrition bias related to missing outcome data: A longitudinal simulation study. Epidemiology, 29, 87–95. https://doi.org/10.1097/EDE.0000000000000755

32.

Liang

Yuan

Wang

Huang

Zeng

Yang

Zhou

(2022). Relationship between perceived stress and depression in Chinese front-line medical staff during COVID-19: A conditional process model. Journal of Affective Disorders, 311, 40–46. https://doi.org/10.1016/j.jad.2022.05.064

33.

Wang

Lin

(2020). Psychological status of medical workforce during the COVID-19 pandemic: A cross-sectional study. Psychiatry Research, 288, 112936. https://doi.org/10.1016/j.psychres.2020.112936

34.

Lüdecke

Ben-Shachar

Patil

Waggoner

Makowski

(2021). Performance: An R package for assessment, comparison and testing of statistical models. Journal of Open Source Software, 6, 3139. https://doi.org/10.21105/joss.03139

35.

Magnúsdóttir

Lovik

Unnarsdóttir

A. B.

McCartney

Ask

Kõiv

Christoffersen

L. A. N.

Johnson

S. U.

Hauksdóttir

Fawns-Ritchie

Helenius

González-Hijón

Ebrahimi

O. V.

Hoffart

Porteous

D. J.

Fang

Jakobsdóttir

Lehto

, . . . Pedersen

O. B.

(2022). Acute COVID-19 severity and mental health morbidity trajectories in patient populations of six nations: An observational study. The Lancet Public Health, 7, e406–e416. https://doi.org/10.1016/S2468-2667(22)00042-1

36.

Marvaldi

Mallet

Dubertret

Moro

M. R.

Guessoum

S. B.

(2021). Anxiety, depression, trauma-related, and sleep disorders among healthcare workers during the COVID-19 pandemic: A systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 126, 252–264. https://doi.org/10.1016/j.neubiorev.2021.03.024

37.

Mazza

M. G.

de Lorenzo

Conte

Poletti

Vai

Bollettini

Melloni

E. M. T.

Furlan

Ciceri

Rovere-Querini

Benedetti

(2020). Anxiety and depression in COVID-19 survivors: Role of inflammatory and clinical predictors. Brain, Behavior, and Immunity, 89, 594–600. https://doi.org/10.1016/j.bbi.2020.07.037

38.

Miller

A. H.

Raison

C. L.

(2016). The role of inflammation in depression: From evolutionary imperative to modern treatment target. Nature Reviews. Immunology, 16, 22–34. https://doi.org/10.1038/nri.2015.5

39.

Moreno

Wykes

Galderisi

Nordentoft

Crossley

Jones

Cannon

Correll

C. U.

Byrne

Carr

Chen

E. Y. H.

Gorwood

Johnson

Kärkkäinen

Krystal

J. H.

Lee

Lieberman

López-Jaramillo

Männikkö

, . . . Arango

(2020). How mental health care should change as a consequence of the COVID-19 pandemic. The Lancet Psychiatry, 7, 813–824. https://doi.org/10.1016/S2215-0366(20)30307-2

40.

Muller

A. E.

Hafstad

E. V.

Himmels

J. P. W.

Smedslund

Flottorp

Stensland

S. Ø.

Stroobants

Van de Velde

Vist

G. E.

(2020). The mental health impact of the covid-19 pandemic on healthcare workers, and interventions to help them: A rapid systematic review. Psychiatry Research, 293, 113441. https://doi.org/10.1016/j.psychres.2020.113441

41.

Nicolaou

Menikou

Lamnisos

Lubenko

Presti

Squatrito

Constantinou

Papacostas

Aydın

Chong

Y. Y.

Chien

W. T.

Cheng

H. Y.

Ruiz

F. J.

Segura-Vargas

M. A.

Garcia-Martin

M. B.

Obando-Posada

D. P.

Vasiliou

V. S.

McHugh

Höfer

, . . . Gloster

A. T.

(2021). Mental health status of healthcare workers during the COVID-19 outbreak. European Journal of Psychology Open, 80, 62–76. https://doi.org/10.1024/2673-8627/a000010

42.

Pancani

Marinucci

Aureli

Riva

(2021). Forced social isolation and mental health: A study on 1,006 Italians under COVID-19 lockdown. Frontiers in Psychology, 12, 663799. https://doi.org/10.3389/fpsyg.2021.663799

43.

Pinheiro

Bates

, . . . R Core Team (2022). Linear and nonlinear mixed effects models [R package nlme version 3.1-157]. Comprehensive R Archive Network (CRAN). https://CRAN.R-project.org/package=nlme

44.

Radloff

L. S.

(1977). The CES-D Scale. Applied Psychological Measurement, 1, 385–401. https://doi.org/10.1177/014662167700100306

45.

Sagherian

Steege

L. M.

Cobb

S. J.

Cho

(2020). Insomnia, fatigue and psychosocial well-being during COVID-19 pandemic: A cross-sectional survey of hospital nursing staff in the United States. Journal of Clinical Nursing. Advance online publication. https://doi.org/10.1111/jocn.15566

46.

Sahu

A. K.

Amrithanand

V. T.

Mathew

Aggarwal

Nayer

Bhoi

(2020). COVID-19 in health care workers – A systematic review and meta-analysis. The American Journal of Emergency Medicine, 38, 1727–1731. https://doi.org/10.1016/j.ajem.2020.05.113

47.

Shafer

A. B.

(2006). Meta-analysis of the factor structures of four depression questionnaires: Beck, CES-D, Hamilton, and Zung. Journal of Clinical Psychology, 62, 123–146. https://doi.org/10.1002/jclp.20213

48.

Spoorthy

M. S.

Pratapa

S. K.

Mahant

(2020). Mental health problems faced by healthcare workers due to the COVID-19 pandemic-A review. Asian Journal of Psychiatry, 51, 102119. https://doi.org/10.1016/j.ajp.2020.102119

49.

Thom

Kuhenert

Born

(2017). 12-Monats-Prävalenz der selbstberichteten ärztlich diagnostizierten Depression in Deutschland [12-month prevalence of self-reported medically diagnosed depression in Germany]. Journal of Health Monitoring, 2, 68–76. https://doi.org/10.17886/RKI-GBE-2017-057

50.

Twisk

J. W. R.

(2006). Applied multilevel analysis: A practical guide (p. 184). Cambridge University Press.

51.

Wagenmakers

E.-J.

Farrell

(2004). AIC model selection using Akaike weights. Psychonomic Bulletin & Review, 11, 192–196. https://doi.org/10.3758/BF03206482

52.

Wasem

Blase

(2023). Human resources development at the hospital since 2000. In Klauber

Wasem

Beivers

Mostert

(Eds.), Hospital report 2023 (1st ed., pp. 3–18). Springer Berlin/Heidelberg.

53.

World Health Organisation. (2020). COVID 19: Public Health Emergency of International Concern (PHEIC). Retrieved June 27, 2022, from https://www.who.int/publications/m/item/covid-19-public-health-emergency-of-international-concern-(pheic)-global-research-and-innovation-forum

54.

World Health Organisation. (2022). Mental Health and COVID-19: Early evidence of the pandemic’s impact: Scientific brief, 2 March 2022. Retrieved June 23, 2022, from https://www.who.int/publications/i/item/WHO-2019-nCoV-Sci_Brief-Mental_health-2022.1

55.

Xiao

Zhang

Kong

Yang

(2020). The effects of social support on sleep quality of medical staff treating patients with coronavirus disease 2019 (COVID-19) in January and February 2020 in China. Medical Science Monitor, 26, e923549. https://doi.org/10.12659/MSM.923549

56.

Xie

Al-Aly

(2022). Risks of mental health outcomes in people with covid-19: Cohort study. BMJ, 376, e068993. https://doi.org/10.1136/bmj-2021-068993

57.

Zeng

Zhao

Y.-M.

Yan

Q.-D.

Liu

S.-Y.

Mei

Yuan

Fan

T.-T.

Yuan

J.-L.

Vitiello

M. V.

Kosten

Kondratiuk

A. L.

Sun

H.-Q.

Tang

X.-D.

Liu

M.-Y.

, . . . Lu

(2022). A systematic review and meta-analysis of long term physical and mental sequelae of COVID-19 pandemic: Call for research priority and action. Molecular Psychiatry, 28(1), 423–433. https://doi.org/10.1038/s41380-022-01614-7

58.

Zhang

W.-R.

Wang

Yin

Zhao

W.-F.

Xue

Peng

Min

B.-Q.

Tian

Leng

H.-X.

J.-L.

Chang

Yang

Shangguan

F.-F.

Yan

T.-Y.

Dong

H.-Q.

Han

Wang

Y.-P.

Cosci

Wang

H.-X.

(2020). Mental health and psychosocial problems of medical health workers during the COVID-19 epidemic in China. Psychotherapy and Psychosomatics, 89, 242–250. https://doi.org/10.1159/000507639

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB