Abstract
Background
The COVID-19 pandemic prompted significant shifts to teleworking, raising questions about potential impacts on employee wellbeing. This study examined the association between self-reported changes to teleworking frequency (relative to before the pandemic) and two indicators of occupational burnout: emotional exhaustion and professionally diagnosed burnout.
Methods
Data were derived from two samples from a digital cohort study based in Geneva, Switzerland: one population-based, and one from a sample of workers who were likely mobilized in the early stages of the COVID-19 pandemic. Emotional exhaustion was measured using the Maslach Burnout Inventory (EE-MBI), while self-reported diagnosed burnout was assessed by asking participants if they had received a professional diagnosis of occupational burnout within the previous 12 months. Participants were categorized based on self-reported telework frequency changes: “no change,” “increase,” “decrease,” “never telework,” and “not possible to telework.” Adjusted regression models for each of the study samples were used to estimate associations between telework changes and burnout outcomes, accounting for sociodemographic, household, and work-related factors.
Results
In the population-based sample of salaried employees (n = 1,332), the median EE-MBI score was 14 (interquartile range: 6–24), and 7.3% reported diagnosed burnout. Compared to those reporting no change in telework frequency (19% of the sample), those reporting a decrease (4%) and those reporting that teleworking was not possible (28.7%) had significantly higher emotional exhaustion scores (adjusted beta (aβ) 5.26 [95% confidence interval: 1.47, 9.04] and aβ 3.51 [0.44, 6.59], respectively) and additionally reported higher odds of diagnosed burnout (adjusted odds ratio (aOR) 10.59 [3.24, 34.57] and aOR 3.42 [1.22, 9.65], respectively). “Increased” (28.9%) and “never” (19.4%) telework statuses were not significantly associated with burnout outcomes. These trends were mirrored in the “mobilized-workers” sample, with the exception that those reporting that teleworking was not possible did not report significantly higher odds of diagnosed burnout compared to those reporting no change in telework frequency.
Conclusions
Decreased teleworking frequency and not having the possibility of telework were associated with higher emotional exhaustion and diagnosed burnout. As organizations reconsider their telework policies in a post-pandemic era, they should consider the impact of such organizational changes on employee wellbeing.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12889-024-21142-z.
Keywords: Diagnosed burnout, Organizational change, Remote work, Emotional exhaustion, Work from home
Background
Occupational burnout is a psychosocial occupational hazard that stems from extended exposure to job-related issues, and exhibits as a state of physical and emotional exhaustion [1]. It affects all job sectors and occupational groups [2] and imposes important personal, social, and economic costs. These include degraded physical and mental health, reduced performance at work, and higher rates of work absenteeism [3]. Most research uses patient-reported scales to measure burnout symptoms, of which emotional exhaustion constitutes a core characteristic [4]. In Switzerland, a meta-analysis of studies published between 2010–2020 estimated the average national prevalence of clinical/severe burnout to be 4% [95% confidence interval: 2, 6], and of severe emotional exhaustion to be 18% [15, 22] [5].
The Job Demands-Resources theory describes a potential pathway to job stress and burnout via work-related “demands” that require sustained mental or physical effort (e.g. organizational change, work overload, interpersonal conflict) [6, 7]. These demands can be mitigated by individual and organizational “resources” (e.g. manager and peer support, autonomy, and time flexibility). In this context, the advent of telework, as well as changing telework policies, can be seen as an important source of organizational change that potentially influences employee mental health [8]. While many definitions of telework exist [9], for this study it is understood to be work done by salaried employees from a location other than the employer’s/client’s premises, for example at home or in co-working spaces. It can act as an organizational “resource”, for example by enabling work-life balance by reducing commuting time or improving employee autonomy through enhanced flexibility, or by reducing emotional exhaustion stemming from workplace-related sustained stimulations or disruptions [10–12]. In certain situations, telework can also act as an organizational “demand”, for example precipitating a perceived loss of social support from colleagues and feelings of isolation, both risk factors for emotional exhaustion and burnout [13].
In the early stages of the COVID-19 pandemic, the teleworking trend was rapidly accelerated as many traditional office environments were closed, and organizations that could technically offer telework were compelled to do so [8, 14]. Many organizations have recently been urging workers to return to the office at least on a partial basis [15], and pre-pandemic [16] and recent [17, 18] literature give some indication that changing policies regarding the availability of telework may differently impact employees, for example due to changes in social interactions with colleagues, adjustments to commuting from home to a physical office space, or disruptions to work-life balance established during periods of prolonged telework. Individual circumstances may also influence how telework or return-to-office policies are perceived (e.g. family and living conditions or whether the role is amenable to telework) [11, 19].
Given the widespread organizational shifts brought about by the COVID-19 pandemic, this study aimed to evaluate the relationship between self-reported changes in telework frequency (compared to before the pandemic) and two indicators of burnout: emotional exhaustion and clinically diagnosed burnout.
Methods
Study design and setting
This study is based on the Specchio-COVID19 digital cohort study (“Specchio” translates as “mirror” in Italian), which was established in Geneva, Switzerland in December 2020 to follow participants of SARS-CoV-2 seroprevalence surveys [20]. Briefly, the cohort consists of individuals recruited after they participated in a population-based seroprevalence survey between April 2020 and June 2022 [21–24]. For these seroprevalence surveys, participants were recruited from a random list of individuals provided by the Geneva cantonal population registry, as well as from a random selection of previous participants of the Bus Santé study, which is a population-based annual repeated cross-sectional study in the Canton of Geneva.
A second non-population-based source of cohort participants consisted of those who participated in a seroprevalence survey (May–September 2020) that targeted workers who were likely to have been mobilized during the initial lockdown of the spring of 2020 [25]. Public and private companies and institutions that were likely operating with on-site activity during the initial lockdown of the spring of 2020 were selected after a consultation process that included representatives from the Geneva cantonal authorities, the Geneva University Hospitals medical directorate, and the University of Geneva. Regardless of the nature of their role, employees of these institutions were then invited, via their human resources or direction departments, to participate in the “mobilized-worker” seroprevalence survey [25].
From November 2020 onwards, participants from either seroprevalence survey who consented to join the Specchio-COVID19 digital cohort were invited to create an online account on the Specchio-COVID19 platform to complete a baseline questionnaire and receive periodic follow-up questionnaires. Additional details on the cohort and each of the surveys can be found in previous literature [20–25].
In November 2022, Specchio-COVID19 digital cohort participants were invited to complete an occupational health questionnaire (hereafter described as “follow-up”) that explored themes related to employment and potential impacts of the COVID-19 pandemic, e.g. on employment conditions, teleworking, job security, workplace interactions, and occupational burnout. English translations of relevant items from the baseline and occupational health questionnaires are included in the supplementary material.
Participants
Specchio-COVID19 participants were eligible to be included in this study if they were of working age (18–64 years) at follow-up, and in salaried employment at baseline. A total of 7,361 were invited to complete the occupational health questionnaire (Fig. 1). Of 3,133 respondents (response rate 42.6%), 2,519 were included for the main analyses; 591 were excluded because they either had not worked during the previous 12 months, were not in salaried employment at follow-up, were not employed before the pandemic, or had completed their baseline questionnaire within the 12 months prior to completing the follow-up questionnaire. A further 3 respondents under 25 years of age were excluded due to low numbers in this age group, 13 respondents were excluded because they reported their sex as “Other” and therefore could not be included in the weighted analysis (the population margins used to apply measurement weights include only frequencies for Male/Female sex), and 7 respondents were excluded because they had missing data in one of the model covariates. The participants were further categorized according to their recruitment source (those recruited from one of the population-based seroprevalence studies were classified as “population-based” (n = 1332), while those recruited from the seroprevalence study of potentially mobilized workers were classified as “mobilized workers” (n = 1245). A small number of participants (n = 58) participated in both studies and are included within each of the analyses that assess these study samples separately.
Fig. 1.
Flowchart of occupational health questionnaire invitations and inclusion into analytical sample
Measures
Sociodemographic and health-related factors were assessed in the baseline questionnaire between December 2020 and June 2022, depending on when the participants joined the cohort. Work-related factors, self-reported changes in telework frequency, emotional exhaustion, and diagnosed burnout were assessed at follow-up, in November 2022. The variables included in the statistical models (exposure, outcomes, and twelve covariates) are listed in Table S1 and are further described below.
Outcomes: Emotional Exhaustion and Diagnosed Burnout
The primary outcome was emotional exhaustion, measured using the French-language Emotional Exhaustion component of the Maslach Burnout Inventory (EE-MBI) [26]. This consists of 9 work-related questions with a 7-point Likert-type response scale (0 = “never”, 6 = “everyday”). These were summed to form one continuous score (range 0–54), with higher scores indicating higher levels of emotional exhaustion. Internal consistency of the scale was high (Cronbach’s α = 0.94). Other outcomes were diagnosed burnout, assessed by asking participants whether they had been professionally diagnosed with occupational burnout within the previous 12 months, and “severe emotional exhaustion”, defined as an EE-MBI score ≥ 30 (a conventionally used cutoff for the French-language version of this scale [27]).
Exposure: Teleworking frequency relative to before the pandemic
The exposure of interest was defined using two questions. The first asked for current telework frequency (every day, at least once per week, irregularly, never, and not possible), and those doing at least some telework were asked whether their telework frequency had changed relative to before the COVID-19 pandemic (no change, decrease, increase). These were used to create our composite exposure variable indicating telework frequency relative to before the pandemic (“no change” [reference group], “decrease”, “increase”, “never”, and “not possible”). An underlying presumption for the difference between the non-teleworking groups is that those who “never” telework might have the option but choose not to, whereas those who indicated it is not possible are limited either by technical challenges or employer restrictions. Explicit information on the technical feasibility or willingness to telework was not available in this study. To reduce potential misclassification between the two non-teleworking groups, free-text responses for job titles were first classified according to the most appropriate 3-digit code of the International Labour Organization’s International Standard Classification of Occupations (ISCO-08) that covers 130 minor groups [28]. By matching these codes with recently developed “teleworkability” indices [14], occupations where teleworking was not technically feasible were reclassified from “never” to “not possible” (n = 105 reclassified).
Model covariates
Sociodemographic and health-related factors
Sociodemographic factors were measured at baseline and included birthdate, sex, and highest educational attainment, categorized as primary (no education or only compulsory schooling), secondary (completion of high school or vocational training), or tertiary (university level [reference group]). Birthdate was used to calculate age at follow-up, categorized as 25–34 [reference group], 35–44, 45–54, and 55–64. Additional measures at baseline included living arrangement ("Couple without children" [reference group], "Couple with children", "Single with children", "With other adults", "Alone"), whether or not the participant had any young children ≤ 10 years old (“No young children” [reference group], “Has young children”), household density (“Not overcrowded” [reference group], and “Overcrowded” if the ratio of inhabitants to bedrooms surpassed 2:1), noise levels at home (“Not noisy” [reference group], “A little noisy”, “Moderately noisy”, “Very noisy”), whether or not the participant has access to a quiet room to work or relax (“Yes” [reference group], “No”), and self-reported general health, dichotomized as “good” [reference group] (comprising “very good” and “good”), and “not good” (“average”, “poor”, and “very poor”).
Work conditions
Using the assigned ISCO-08 codes, occupation was defined using the sub-major groups, further grouping “associate” levels together with their linked group (e.g. grouping “Business and Administration Associate Professionals” with “Business and Administration Professionals”). The largest group was chosen as the reference group (“Business and Administration Professionals”). A variable indicating whether participants reported any pandemic-related work changes (job loss, resignation, new job, change in position or career, new work schedule) was also included (reference group: “No change”). Work contracts were quantified in percentage terms, with 100% equivalent to full-time employment. By comparing the total percentage at baseline with that at follow-up, changes were assessed as “no change” [reference group], “reduced”, and “increased”.
Other descriptive variables
Accounting for household composition and with reference to income information from the Geneva population [29], household income was categorized as low (below the first quartile), middle (between the first and third quartiles), or high (above the third quartile). Work social support was evaluated using a French-language version of the Karasek Job Content Questionnaire [30, 31]. The remaining variables listed in the descriptive tables were assessed using single-item questions.
Statistical analyses
Descriptive analyses
Separately for each sample from the two population sources, variables were described using the frequency and percentage for categorical variables and the median and interquartile range (IQR) for continuous variables. Associations with the exposure or outcomes were examined using the Pearson’s chi-squared and Kruskal–Wallis rank sum tests. Characteristics of respondents and non-respondents were compared descriptively for the total cohort participants.
Regression analyses
Linear and logistic regression models were used to test the associations between changing telework frequency and three outcomes (EE-MBI score, severe emotional exhaustion, and diagnosed burnout). Twelve covariates (identified using a directed acyclic graph (DAG), Figure S1) are included in the full models: age group, sex, education level, living arrangement, having young children, household density, noise levels at home, access to a quiet room, self-reported general health, occupation group, pandemic-related work changes, and changes to contract percentages. No strong associations or multicollinearity were found between any of these variables (Cramér’s V < 0.5 and GVIF1/(2×df) < 2). Other socio-demographic and work-related factors were measured (e.g. work social support, overtime work, feeling overworked), but based on the DAG these were considered as being in the causal pathway and were not included in the models. No evidence of an effect of interaction between sex and teleworking group was found for the continuous EE-MBI score (p > 0.90), severe emotional exhaustion (p = 0.70), or diagnosed burnout (p = 0.80).
For each outcome, unadjusted estimates and 95% confidence intervals (CIs) were calculated using univariable models for the exposure of interest (teleworking frequency relative to before the pandemic; reference group: “no change”). Adjusted regression coefficients (aβ), adjusted odds ratios (aOR), and their 95% CIs were calculated from multivariable linear (EE-MBI score) or logistic (severe emotional exhaustion and diagnosed burnout) regression models, respectively, that included the exposure of interest and the twelve previously described covariates.
The regression estimates were obtained separately for each of the samples obtained from the two population sources. For the “population-based” sample only, the regression analyses incorporated survey sampling weights (using population margins for age, sex, and education obtained from the Cantonal Statistical Office) to generate estimates that are representative for the canton of Geneva. Estimates from each analysis are presented only for the exposure variable, as the multivariable models allow only for the interpretation of adjusted estimates from the exposure variable [32].
Missing data
As few participants had missing data in at least one of the covariates included in the models (n = 7, all in the “overcrowded” covariate), these participants were excluded and we performed complete-case analyses.
Sensitivity analyses
The robustness of our results was tested through two sensitivity analyses. First, the regression models were repeated keeping the “Not possible” group as the reference category of the teleworking exposure variable, to allow direct comparison against those reporting that they “never” telework. Second, to check if the associations within the groups reporting changes to telework frequency were not simply due to them also reporting changes in their contractual hours or pandemic-related work changes, the regressions were repeated using a restricted sample that excluded those who reported any change in their contract percentage (relative to baseline) or who reported any pandemic-related work change. These sensitivity analyses were run for each sample from the two recruitment sources.
All statistical analyses were performed using R, version 4.4.1 (https://www.R-project.org/), using the tidyverse (2.0.0), survey (4.4.2), srvyr (1.3.0), gtsummary (2.0.3), flextable (0.9.6), and forestplot (3.1.5) packages. All code files are publicly accessible on GitHub (https://github.com/UEP-HUG/Telework-Burnout-Public).
Results
Comparison of respondents vs. non-respondents
Baseline characteristics of respondents (n = 3,133) and non-respondents (n = 4,228) are summarized in Table S2. Respondents had a higher representation of female sex and of older age groups, as well as lower representation of low household income.
Descriptive summaries
Compared to the professionally active population of the Canton of Geneva, the included participants from both the “population-based” (n = 1,332) and “mobilized-workers” (n = 1,245) samples tended to be older, more highly educated, and had a higher representation of females (results not shown).
Unless otherwise indicated, the below summaries of the descriptive analyses apply to both of the samples. Relative to teleworkers, non-teleworkers tended to have lower education, household income, and financial security, and also less frequently reported that they work overtime. Teleworkers reporting a decrease in telework frequency less frequently reported people management as a main duty, more frequently reported having experienced aggression from colleagues, and more frequently reporting feeling overworked. In the population-representative sample, teleworkers reporting an increase in telework frequency were more likely to have young children (Table 1 and Table 2).
Table 1.
Unweighted prevalence of various participant characteristics according to teleworking group (sample source: Population-based)
| Teleworking group | ||||||||
|---|---|---|---|---|---|---|---|---|
| Characteristic | N | Overall N = 1332 |
No change N = 253 |
Decrease N = 53 |
Increase N = 385 |
Not possible N = 382 |
Never N = 259 |
p-valuea |
| Age group | 1,332 | 0.118 | ||||||
| 25–34 | 81 (6.1%) | 10 (4.0%) | 2 (3.8%) | 31 (8.1%) | 30 (7.9%) | 8 (3.1%) | ||
| 35–44 | 283 (21%) | 53 (21%) | 11 (21%) | 91 (24%) | 72 (19%) | 56 (22%) | ||
| 45–54 | 513 (39%) | 103 (41%) | 18 (34%) | 150 (39%) | 141 (37%) | 101 (39%) | ||
| 55–64 | 455 (34%) | 87 (34%) | 22 (42%) | 113 (29%) | 139 (36%) | 94 (36%) | ||
| Sex | 1,332 | 0.103 | ||||||
| Male | 554 (42%) | 123 (49%) | 24 (45%) | 159 (41%) | 149 (39%) | 99 (38%) | ||
| Female | 778 (58%) | 130 (51%) | 29 (55%) | 226 (59%) | 233 (61%) | 160 (62%) | ||
| Education | 1,332 | < 0.001 | ||||||
| Tertiary | 910 (68%) | 193 (76%) | 34 (64%) | 299 (78%) | 215 (56%) | 169 (65%) | ||
| Secondary | 395 (30%) | 59 (23%) | 18 (34%) | 86 (22%) | 146 (38%) | 86 (33%) | ||
| Primary | 27 (2.0%) | 1 (0.4%) | 1 (1.9%) | 0 (0%) | 21 (5.5%) | 4 (1.5%) | ||
| Household income | 1,332 | < 0.001 | ||||||
| High | 230 (17%) | 51 (20%) | 13 (25%) | 93 (24%) | 21 (5.5%) | 52 (20%) | ||
| Middle | 777 (58%) | 159 (63%) | 30 (57%) | 221 (57%) | 226 (59%) | 141 (54%) | ||
| Low | 139 (10%) | 17 (6.7%) | 4 (7.5%) | 22 (5.7%) | 66 (17%) | 30 (12%) | ||
| Don't want to answer | 186 (14%) | 26 (10%) | 6 (11%) | 49 (13%) | 69 (18%) | 36 (14%) | ||
| Financial security | 1,332 | < 0.001 | ||||||
| Can cover expenses | 575 (43%) | 109 (43%) | 23 (43%) | 150 (39%) | 177 (46%) | 116 (45%) | ||
| Cautious / struggling | 166 (12%) | 22 (8.7%) | 6 (11%) | 31 (8.1%) | 77 (20%) | 30 (12%) | ||
| Comfortable | 531 (40%) | 112 (44%) | 22 (42%) | 189 (49%) | 106 (28%) | 102 (39%) | ||
| Don't want to answer | 60 (4.5%) | 10 (4.0%) | 2 (3.8%) | 15 (3.9%) | 22 (5.8%) | 11 (4.2%) | ||
| Ethnicity | 1,321 | 0.610 | ||||||
| European-Caucasian | 1,251 (95%) | 240 (96%) | 48 (92%) | 359 (94%) | 355 (94%) | 249 (96%) | ||
| Other | 70 (5.3%) | 11 (4.4%) | 4 (7.7%) | 23 (6.0%) | 22 (5.8%) | 10 (3.9%) | ||
| Living arrangement | 1,332 | 0.805 | ||||||
| Couple without children | 239 (18%) | 42 (17%) | 14 (26%) | 64 (17%) | 73 (19%) | 46 (18%) | ||
| Couple with children | 758 (57%) | 152 (60%) | 30 (57%) | 217 (56%) | 211 (55%) | 148 (57%) | ||
| Single with children | 115 (8.6%) | 20 (7.9%) | 3 (5.7%) | 38 (9.9%) | 37 (9.7%) | 17 (6.6%) | ||
| With other adults | 43 (3.2%) | 10 (4.0%) | 0 (0%) | 11 (2.9%) | 13 (3.4%) | 9 (3.5%) | ||
| Alone | 177 (13%) | 29 (11%) | 6 (11%) | 55 (14%) | 48 (13%) | 39 (15%) | ||
| Has young children | 1,332 | 293 (22%) | 59 (23%) | 9 (17%) | 109 (28%) | 72 (19%) | 44 (17%) | 0.003 |
| Overcrowded household | 1,332 | 18 (1.4%) | 5 (2.0%) | 0 (0%) | 5 (1.3%) | 2 (0.5%) | 6 (2.3%) | 0.263 |
| Noise level at home | 1,332 | 0.231 | ||||||
| Not noisy | 361 (27%) | 75 (30%) | 15 (28%) | 109 (28%) | 90 (24%) | 72 (28%) | ||
| A little noisy | 568 (43%) | 110 (43%) | 23 (43%) | 154 (40%) | 175 (46%) | 106 (41%) | ||
| Moderately noisy | 332 (25%) | 50 (20%) | 10 (19%) | 105 (27%) | 95 (25%) | 72 (28%) | ||
| Very noisy | 71 (5.3%) | 18 (7.1%) | 5 (9.4%) | 17 (4.4%) | 22 (5.8%) | 9 (3.5%) | ||
| No access to a quiet room | 1,332 | 234 (18%) | 40 (16%) | 11 (21%) | 63 (16%) | 80 (21%) | 40 (15%) | 0.276 |
| General health: not good | 1,332 | 101 (7.6%) | 18 (7.1%) | 3 (5.7%) | 24 (6.2%) | 38 (9.9%) | 18 (6.9%) | 0.336 |
| Mental health: not good | 1,332 | 142 (11%) | 24 (9.5%) | 6 (11%) | 44 (11%) | 42 (11%) | 26 (10%) | 0.941 |
| Chronic condition | 1,332 | 289 (22%) | 50 (20%) | 17 (32%) | 71 (18%) | 88 (23%) | 63 (24%) | 0.100 |
| Teleworking frequency | 1,332 | < 0.001 | ||||||
| Occasionally (irregular) | 272 (20%) | 113 (45%) | 22 (42%) | 137 (36%) | 0 (0%) | 0 (0%) | ||
| 1 + days/week | 389 (29%) | 125 (49%) | 29 (55%) | 235 (61%) | 0 (0%) | 0 (0%) | ||
| Every day | 30 (2.3%) | 15 (5.9%) | 2 (3.8%) | 13 (3.4%) | 0 (0%) | 0 (0%) | ||
| Never | 641 (48%) | 0 (0%) | 0 (0%) | 0 (0%) | 382 (100%) | 259 (100%) | ||
| Experienced work changes | 1,332 | 77 (5.8%) | 9 (3.6%) | 3 (5.7%) | 32 (8.3%) | 23 (6.0%) | 10 (3.9%) | 0.071 |
| Change in contracted hoursb | 1,332 | 0.740 | ||||||
| No change | 982 (74%) | 194 (77%) | 39 (74%) | 292 (76%) | 272 (71%) | 185 (71%) | ||
| Increased | 245 (18%) | 42 (17%) | 11 (21%) | 67 (17%) | 74 (19%) | 51 (20%) | ||
| Reduced | 105 (7.9%) | 17 (6.7%) | 3 (5.7%) | 26 (6.8%) | 36 (9.4%) | 23 (8.9%) | ||
| Management duties | 1,332 | 0.005 | ||||||
| None | 696 (52%) | 121 (48%) | 27 (51%) | 182 (47%) | 223 (58%) | 143 (55%) | ||
| Some duties | 389 (29%) | 85 (34%) | 21 (40%) | 115 (30%) | 104 (27%) | 64 (25%) | ||
| Main duty | 247 (19%) | 47 (19%) | 5 (9.4%) | 88 (23%) | 55 (14%) | 52 (20%) | ||
| Work social support | 1,332 | 0.010 | ||||||
| High | 957 (72%) | 196 (77%) | 36 (68%) | 278 (72%) | 250 (65%) | 197 (76%) | ||
| Medium | 353 (27%) | 55 (22%) | 14 (26%) | 102 (26%) | 124 (32%) | 58 (22%) | ||
| Low | 22 (1.7%) | 2 (0.8%) | 3 (5.7%) | 5 (1.3%) | 8 (2.1%) | 4 (1.5%) | ||
| Aggression from colleagues | 1,332 | 118 (8.9%) | 16 (6.3%) | 9 (17%) | 32 (8.3%) | 48 (13%) | 13 (5.0%) | 0.001 |
| Fears losing job | 1,332 | 159 (12%) | 30 (12%) | 9 (17%) | 56 (15%) | 39 (10%) | 25 (9.7%) | 0.188 |
| Feels overworked | 1,332 | 475 (36%) | 84 (33%) | 26 (49%) | 154 (40%) | 132 (35%) | 79 (31%) | 0.025 |
| Works overtime | 1,332 | < 0.001 | ||||||
| Rarely/Never | 256 (19%) | 31 (12%) | 8 (15%) | 44 (11%) | 110 (29%) | 63 (24%) | ||
| Occasionally | 499 (37%) | 104 (41%) | 17 (32%) | 149 (39%) | 140 (37%) | 89 (34%) | ||
| Often | 577 (43%) | 118 (47%) | 28 (53%) | 192 (50%) | 132 (35%) | 107 (41%) | ||
|
EE-MBI score, median (IQR) [Weighted estimate]c |
1,332 |
14 (7, 25) [14 (6, 24)] |
13 (6, 22) [13 (7, 24)] |
18 (10, 29) [18 (9, 29)] |
16 (9, 25) [16 (10, 25)] |
14 (6, 27) [12 (4, 23)] |
13 (6, 24) [14 (6, 22)] |
0.002 |
|
Severe emotional exhaustion [Weighted estimate]c |
1,332 |
236 (18%) [16.9%] |
37 (15%) [16.7%] |
13 (25%) [22.6%] |
70 (18%) [17.5%] |
78 (20%) [16.9%] |
38 (15%) [15.1%] |
0.138 |
|
Diagnosed burnout [Weighted estimate]c |
1,332 |
91 (6.8%) [7.3%] |
11 (4.3%) [3.5%] |
9 (17%) [18.9%] |
27 (7.0%) [6.6%] |
28 (7.3%) [9.5%] |
16 (6.2%) [4.4%] |
0.023 |
Data are n (%) unless otherwise stated. % may not add up to 100% due to rounding
aKruskal–Wallis rank sum test; Pearson's Chi-squared test
bRelated to the pandemic (e.g. career change, promotion)
cWeighted estimate after incorporating survey sampling weights (for age, sex, and education, from the resident professionally active population of the canton of Geneva)
Table 2.
Unweighted prevalence of various participant characteristics according to teleworking group (sample source: Mobilized workers)
| Teleworking group | ||||||||
|---|---|---|---|---|---|---|---|---|
| Characteristic | N | Overall N = 1245 |
No change N = 177 |
Decrease N = 62 |
Increase N = 261 |
Not possible N = 572 |
Never N = 173 |
p-valuea |
| Age group | 1,245 | 0.046 | ||||||
| 25–34 | 107 (8.6%) | 9 (5.1%) | 7 (11%) | 21 (8.0%) | 61 (11%) | 9 (5.2%) | ||
| 35–44 | 283 (23%) | 36 (20%) | 21 (34%) | 61 (23%) | 135 (24%) | 30 (17%) | ||
| 45–54 | 489 (39%) | 71 (40%) | 19 (31%) | 108 (41%) | 215 (38%) | 76 (44%) | ||
| 55–64 | 366 (29%) | 61 (34%) | 15 (24%) | 71 (27%) | 161 (28%) | 58 (34%) | ||
| Sex | 1,245 | 0.143 | ||||||
| Male | 485 (39%) | 78 (44%) | 21 (34%) | 94 (36%) | 214 (37%) | 78 (45%) | ||
| Female | 760 (61%) | 99 (56%) | 41 (66%) | 167 (64%) | 358 (63%) | 95 (55%) | ||
| Education | 1,245 | < 0.001 | ||||||
| Tertiary | 799 (64%) | 133 (75%) | 48 (77%) | 204 (78%) | 325 (57%) | 89 (51%) | ||
| Secondary | 419 (34%) | 40 (23%) | 12 (19%) | 54 (21%) | 231 (40%) | 82 (47%) | ||
| Primary | 27 (2.2%) | 4 (2.3%) | 2 (3.2%) | 3 (1.1%) | 16 (2.8%) | 2 (1.2%) | ||
| Household income | 1,245 | < 0.001 | ||||||
| High | 161 (13%) | 37 (21%) | 17 (27%) | 53 (20%) | 30 (5.2%) | 24 (14%) | ||
| Middle | 711 (57%) | 103 (58%) | 30 (48%) | 147 (56%) | 329 (58%) | 102 (59%) | ||
| Low | 179 (14%) | 10 (5.6%) | 6 (9.7%) | 22 (8.4%) | 124 (22%) | 17 (9.8%) | ||
| Don't want to answer | 194 (16%) | 27 (15%) | 9 (15%) | 39 (15%) | 89 (16%) | 30 (17%) | ||
| Financial security | 1,245 | < 0.001 | ||||||
| Can cover expenses | 580 (47%) | 66 (37%) | 24 (39%) | 121 (46%) | 277 (48%) | 92 (53%) | ||
| Cautious / struggling | 157 (13%) | 19 (11%) | 7 (11%) | 18 (6.9%) | 94 (16%) | 19 (11%) | ||
| Comfortable | 450 (36%) | 82 (46%) | 29 (47%) | 112 (43%) | 173 (30%) | 54 (31%) | ||
| Don't want to answer | 58 (4.7%) | 10 (5.6%) | 2 (3.2%) | 10 (3.8%) | 28 (4.9%) | 8 (4.6%) | ||
| Ethnicity | 1,234 | 0.524 | ||||||
| European-Caucasian | 1,182 (96%) | 165 (95%) | 59 (98%) | 249 (96%) | 543 (95%) | 166 (98%) | ||
| Other | 52 (4.2%) | 9 (5.2%) | 1 (1.7%) | 11 (4.2%) | 27 (4.7%) | 4 (2.4%) | ||
| Living arrangement | 1,245 | 0.415 | ||||||
| Couple without children | 279 (22%) | 48 (27%) | 16 (26%) | 61 (23%) | 119 (21%) | 35 (20%) | ||
| Couple with children | 615 (49%) | 85 (48%) | 37 (60%) | 125 (48%) | 279 (49%) | 89 (51%) | ||
| Single with children | 120 (9.6%) | 18 (10%) | 2 (3.2%) | 28 (11%) | 55 (9.6%) | 17 (9.8%) | ||
| With other adults | 38 (3.1%) | 5 (2.8%) | 2 (3.2%) | 4 (1.5%) | 23 (4.0%) | 4 (2.3%) | ||
| Alone | 193 (16%) | 21 (12%) | 5 (8.1%) | 43 (16%) | 96 (17%) | 28 (16%) | ||
| Has young children | 1,245 | 243 (20%) | 32 (18%) | 16 (26%) | 53 (20%) | 112 (20%) | 30 (17%) | 0.659 |
| Overcrowded household | 1,245 | 14 (1.1%) | 0 (0%) | 3 (4.8%) | 2 (0.8%) | 5 (0.9%) | 4 (2.3%) | 0.014 |
| Noise level at home | 1,245 | 0.632 | ||||||
| Not noisy | 425 (34%) | 54 (31%) | 24 (39%) | 92 (35%) | 201 (35%) | 54 (31%) | ||
| A little noisy | 511 (41%) | 74 (42%) | 25 (40%) | 110 (42%) | 238 (42%) | 64 (37%) | ||
| Moderately noisy | 258 (21%) | 42 (24%) | 10 (16%) | 48 (18%) | 114 (20%) | 44 (25%) | ||
| Very noisy | 51 (4.1%) | 7 (4.0%) | 3 (4.8%) | 11 (4.2%) | 19 (3.3%) | 11 (6.4%) | ||
| No access to a quiet room | 1,245 | 241 (19%) | 26 (15%) | 19 (31%) | 48 (18%) | 111 (19%) | 37 (21%) | 0.086 |
| General health: not good | 1,245 | 85 (6.8%) | 9 (5.1%) | 2 (3.2%) | 21 (8.0%) | 44 (7.7%) | 9 (5.2%) | 0.391 |
| Mental health: not good | 1,245 | 117 (9.4%) | 16 (9.0%) | 7 (11%) | 23 (8.8%) | 49 (8.6%) | 22 (13%) | 0.542 |
| Chronic condition | 1,245 | 273 (22%) | 37 (21%) | 12 (19%) | 55 (21%) | 135 (24%) | 34 (20%) | 0.751 |
| Teleworking frequency | 1,245 | < 0.001 | ||||||
| Occasionally (irregular) | 256 (21%) | 99 (56%) | 31 (50%) | 126 (48%) | 0 (0%) | 0 (0%) | ||
| 1 + days/week | 236 (19%) | 77 (44%) | 31 (50%) | 128 (49%) | 0 (0%) | 0 (0%) | ||
| Every day | 8 (0.6%) | 1 (0.6%) | 0 (0%) | 7 (2.7%) | 0 (0%) | 0 (0%) | ||
| Never | 745 (60%) | 0 (0%) | 0 (0%) | 0 (0%) | 572 (100%) | 173 (100%) | ||
| Experienced work changes | 1,245 | 75 (6.0%) | 6 (3.4%) | 2 (3.2%) | 18 (6.9%) | 40 (7.0%) | 9 (5.2%) | 0.339 |
| Change in contracted hoursb | 1,245 | 0.180 | ||||||
| No change | 898 (72%) | 141 (80%) | 47 (76%) | 182 (70%) | 404 (71%) | 124 (72%) | ||
| Increased | 255 (20%) | 30 (17%) | 9 (15%) | 62 (24%) | 118 (21%) | 36 (21%) | ||
| Reduced | 92 (7.4%) | 6 (3.4%) | 6 (9.7%) | 17 (6.5%) | 50 (8.7%) | 13 (7.5%) | ||
| Management duties | 1,245 | < 0.001 | ||||||
| None | 567 (46%) | 64 (36%) | 34 (55%) | 119 (46%) | 270 (47%) | 80 (46%) | ||
| Some duties | 379 (30%) | 60 (34%) | 13 (21%) | 72 (28%) | 195 (34%) | 39 (23%) | ||
| Main duty | 299 (24%) | 53 (30%) | 15 (24%) | 70 (27%) | 107 (19%) | 54 (31%) | ||
| Work social support | 1,245 | 0.953 | ||||||
| High | 874 (70%) | 129 (73%) | 42 (68%) | 187 (72%) | 393 (69%) | 123 (71%) | ||
| Medium | 346 (28%) | 44 (25%) | 19 (31%) | 69 (26%) | 166 (29%) | 48 (28%) | ||
| Low | 25 (2.0%) | 4 (2.3%) | 1 (1.6%) | 5 (1.9%) | 13 (2.3%) | 2 (1.2%) | ||
| Aggression from colleagues | 1,245 | 129 (10%) | 11 (6.2%) | 12 (19%) | 30 (11%) | 58 (10%) | 18 (10%) | 0.059 |
| Fears losing job | 1,245 | 134 (11%) | 11 (6.2%) | 8 (13%) | 28 (11%) | 68 (12%) | 19 (11%) | 0.301 |
| Feels overworked | 1,245 | 526 (42%) | 68 (38%) | 29 (47%) | 114 (44%) | 245 (43%) | 70 (40%) | 0.716 |
| Works overtime | 1,245 | < 0.001 | ||||||
| Rarely/Never | 246 (20%) | 23 (13%) | 4 (6.5%) | 28 (11%) | 158 (28%) | 33 (19%) | ||
| Occasionally | 464 (37%) | 56 (32%) | 28 (45%) | 99 (38%) | 214 (37%) | 67 (39%) | ||
| Often | 535 (43%) | 98 (55%) | 30 (48%) | 134 (51%) | 200 (35%) | 73 (42%) | ||
| EE-MBI score, median (IQR) | 1,245 | 16 (8, 26) | 13 (8, 23) | 20 (12, 30) | 16 (9, 26) | 17 (8, 28) | 12 (6, 25) | 0.010 |
| Severe emotional exhaustion | 1,245 | 245 (20%) | 25 (14%) | 16 (26%) | 47 (18%) | 126 (22%) | 31 (18%) | 0.102 |
| Diagnosed burnout | 1,245 | 82 (6.6%) | 9 (5.1%) | 9 (15%) | 14 (5.4%) | 40 (7.0%) | 10 (5.8%) | 0.093 |
Data are n (%) unless otherwise stated. % may not add up to 100% due to rounding
aKruskal–Wallis rank sum test; Pearson's Chi-squared test
bRelated to the pandemic (e.g. career change, promotion)
Overall, the burnout measures were similar for the two samples (Table 1 and Table 2), though the EE-MBI score was slightly higher (p = 0.024) in the mobilized-workers sample (median: 16) than in the population-based sample (median: 14). In both samples, female participants on average reported higher EE-MBI scores and higher frequencies of severe emotional exhaustion and diagnosed burnout (Table 3 and Table 4). Overall, the frequency of diagnosed burnout tended to increase with increasing EE-MBI scores (Figure S2). Other variables significantly associated with emotional exhaustion and diagnosed burnout included access to a quiet room, self-reported health, self-reported mental state, chronic condition, pandemic-related work changes, social support in the workplace, aggression from colleagues, feeling overworked, working overtime, and job security (Table 3 and Table 4). In the population-based sample only, living arrangement and financial security were also associated with emotional exhaustion and diagnosed burnout.
Table 3.
Participant characteristics by outcomes (sample source: Population-based). Estimates are unweighted
| EE-MBI | Severe emotional exhaustion | Diagnosed burnout | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | N | Score median (IQR) |
p-valuea | No N = 1096b |
Yes N = 236b |
p-valuec | No N = 1241b |
Yes N = 91b |
p-valuec |
| Teleworking group (n = 1332) | 0.012 | 0.138 | 0.023 | ||||||
| No change | 253 | 13 (6.0–22.0) | 216 (85%) | 37 (15%) | 242 (96%) | 11 (4.3%) | |||
| Decrease | 53 | 18 (10.0–29.0) | 40 (75%) | 13 (25%) | 44 (83%) | 9 (17%) | |||
| Increase | 385 | 16 (9.0–25.0) | 315 (82%) | 70 (18%) | 358 (93%) | 27 (7.0%) | |||
| Not possible | 382 | 14 (6.0–27.0) | 304 (80%) | 78 (20%) | 354 (93%) | 28 (7.3%) | |||
| Never | 259 | 13 (6.0–24.0) | 221 (85%) | 38 (15%) | 243 (94%) | 16 (6.2%) | |||
| Age group (n = 1332) | < 0.001 | 0.125 | 0.888 | ||||||
| 25–34 | 81 | 14 (9.0–22.0) | 70 (86%) | 11 (14%) | 77 (95%) | 4 (4.9%) | |||
| 35–44 | 283 | 17 (9.5–27.0) | 220 (78%) | 63 (22%) | 263 (93%) | 20 (7.1%) | |||
| 45–54 | 513 | 14 (7.0–25.0) | 425 (83%) | 88 (17%) | 476 (93%) | 37 (7.2%) | |||
| 55–64 | 455 | 12 (5.0–24.0) | 381 (84%) | 74 (16%) | 425 (93%) | 30 (6.6%) | |||
| Sex (n = 1332) | < 0.001 | < 0.001 | 0.003 | ||||||
| Male | 554 | 13 (6.0–22.8) | 483 (87%) | 71 (13%) | 530 (96%) | 24 (4.3%) | |||
| Female | 778 | 15 (8.0–27.0) | 613 (79%) | 165 (21%) | 711 (91%) | 67 (8.6%) | |||
| Education (n = 1332) | 0.247 | 0.456 | 0.391 | ||||||
| Tertiary | 910 | 14 (7.0–25.0) | 753 (83%) | 157 (17%) | 853 (94%) | 57 (6.3%) | |||
| Secondary | 395 | 14 (6.0–26.5) | 319 (81%) | 76 (19%) | 364 (92%) | 31 (7.8%) | |||
| Primary | 27 | 10 (3.5–18.0) | 24 (89%) | 3 (11%) | 24 (89%) | 3 (11%) | |||
| Household income (n = 1332) | 0.549 | 0.531 | 0.086 | ||||||
| High | 230 | 13.5 (7.0–22.8) | 197 (86%) | 33 (14%) | 221 (96%) | 9 (3.9%) | |||
| Middle | 777 | 14 (7.0–26.0) | 634 (82%) | 143 (18%) | 717 (92%) | 60 (7.7%) | |||
| Low | 139 | 15 (6.5–25.0) | 114 (82%) | 25 (18%) | 126 (91%) | 13 (9.4%) | |||
| Don't want to answer | 186 | 13 (6.0–26.0) | 151 (81%) | 35 (19%) | 177 (95%) | 9 (4.8%) | |||
| Financial security (n = 1332) | < 0.001 | < 0.001 | 0.015 | ||||||
| Can cover expenses | 575 | 16 (8.0–26.5) | 464 (81%) | 111 (19%) | 532 (93%) | 43 (7.5%) | |||
| Cautious / struggling | 166 | 18 (9.0–30.0) | 122 (73%) | 44 (27%) | 147 (89%) | 19 (11%) | |||
| Comfortable | 531 | 12 (6.0–21.5) | 462 (87%) | 69 (13%) | 503 (95%) | 28 (5.3%) | |||
| Don't want to answer | 60 | 14.5 (5.5–27.0) | 48 (80%) | 12 (20%) | 59 (98%) | 1 (1.7%) | |||
| Ethnicity (n = 1321) | 0.05 | 0.022 | > 0.999 | ||||||
| European-Caucasian | 1,251 | 14 (7.0–25.0) | 1,037 (83%) | 214 (17%) | 1,166 (93%) | 85 (6.8%) | |||
| Other | 70 | 16 (8.2–30.0) | 50 (71%) | 20 (29%) | 65 (93%) | 5 (7.1%) | |||
| Living arrangement (n = 1332) | 0.003 | 0.053 | 0.032 | ||||||
| Couple without children | 239 | 12 (6.0–23.0) | 203 (85%) | 36 (15%) | 231 (97%) | 8 (3.3%) | |||
| Couple with children | 758 | 14 (7.0–24.0) | 635 (84%) | 123 (16%) | 708 (93%) | 50 (6.6%) | |||
| Single with children | 115 | 20 (10.0–27.5) | 88 (77%) | 27 (23%) | 102 (89%) | 13 (11%) | |||
| With other adults | 43 | 16 (7.5–28.5) | 35 (81%) | 8 (19%) | 38 (88%) | 5 (12%) | |||
| Alone | 177 | 15 (8.0–26.0) | 135 (76%) | 42 (24%) | 162 (92%) | 15 (8.5%) | |||
| Has young children (n = 1332) | 0.014 | 0.067 | > 0.999 | ||||||
| No young children | 1,039 | 14 (7.0–25.0) | 866 (83%) | 173 (17%) | 968 (93%) | 71 (6.8%) | |||
| Has young children | 293 | 16 (9.0–27.0) | 230 (78%) | 63 (22%) | 273 (93%) | 20 (6.8%) | |||
| Overcrowded household (n = 1332) | 0.885 | 0.415 | 0.232 | ||||||
| Not overcrowded | 1,314 | 14 (7.0–25.0) | 1,083 (82%) | 231 (18%) | 1,226 (93%) | 88 (6.7%) | |||
| Overcrowded | 18 | 12.5 (5.2–29.8) | 13 (72%) | 5 (28%) | 15 (83%) | 3 (17%) | |||
| Noise level at home (n = 1332) | < 0.001 | 0.699 | 0.642 | ||||||
| Not noisy | 361 | 13 (6.0–24.0) | 302 (84%) | 59 (16%) | 333 (92%) | 28 (7.8%) | |||
| A little noisy | 568 | 13 (6.0–25.0) | 469 (83%) | 99 (17%) | 535 (94%) | 33 (5.8%) | |||
| Moderately noisy | 332 | 17 (8.0–26.0) | 269 (81%) | 63 (19%) | 307 (92%) | 25 (7.5%) | |||
| Very noisy | 71 | 19 (13.5–27.0) | 56 (79%) | 15 (21%) | 66 (93%) | 5 (7.0%) | |||
| Access to a quiet room (n = 1332) | 0.021 | 0.844 | 0.032 | ||||||
| Yes | 1,098 | 14 (6.0–25.0) | 905 (82%) | 193 (18%) | 1,031 (94%) | 67 (6.1%) | |||
| No | 234 | 18 (9.0–26.0) | 191 (82%) | 43 (18%) | 210 (90%) | 24 (10%) | |||
| Self-reported general health (n = 1332) | < 0.001 | < 0.001 | 0.059 | ||||||
| Good | 1,231 | 14 (7.0–24.0) | 1,028 (84%) | 203 (16%) | 1,152 (94%) | 79 (6.4%) | |||
| Not good | 101 | 24 (14.0–34.0) | 68 (67%) | 33 (33%) | 89 (88%) | 12 (12%) | |||
| Self-reported mental state (n = 1332) | < 0.001 | < 0.001 | 0.041 | ||||||
| Good | 1,190 | 13 (6.0–24.0) | 1,014 (85%) | 176 (15%) | 1,115 (94%) | 75 (6.3%) | |||
| Not good | 142 | 26.5 (15.2–34.0) | 82 (58%) | 60 (42%) | 126 (89%) | 16 (11%) | |||
| Chronic condition (n = 1332) | < 0.001 | < 0.001 | 0.010 | ||||||
| No | 1,043 | 13 (6.0–24.0) | 878 (84%) | 165 (16%) | 982 (94%) | 61 (5.8%) | |||
| Yes | 289 | 18 (9.0–29.0) | 218 (75%) | 71 (25%) | 259 (90%) | 30 (10%) | |||
| Teleworking frequency (n = 1332) | 0.073 | 0.685 | 0.337 | ||||||
| Occasionally (irregular) | 272 | 13 (7.0–25.0) | 226 (83%) | 46 (17%) | 256 (94%) | 16 (5.9%) | |||
| 1 + days/week | 389 | 16 (9.0–25.0) | 318 (82%) | 71 (18%) | 358 (92%) | 31 (8.0%) | |||
| Every day | 30 | 10 (3.0–17.8) | 27 (90%) | 3 (10%) | 30 (100%) | 0 (0%) | |||
| Never | 641 | 14 (6.0–25.0) | 525 (82%) | 116 (18%) | 597 (93%) | 44 (6.9%) | |||
| Pandemic-related work change (n = 1332) | 0.033 | 0.236 | > 0.999 | ||||||
| No change | 1,255 | 14 (7.0–25.0) | 1,037 (83%) | 218 (17%) | 1,169 (93%) | 86 (6.9%) | |||
| Change | 77 | 19 (11.0–29.0) | 59 (77%) | 18 (23%) | 72 (94%) | 5 (6.5%) | |||
| Change in contracted hours (n = 1332) | 0.266 | 0.276 | 0.564 | ||||||
| No change | 982 | 14 (7.0–25.0) | 815 (83%) | 167 (17%) | 911 (93%) | 71 (7.2%) | |||
| Increased | 245 | 15 (8.0–27.0) | 193 (79%) | 52 (21%) | 230 (94%) | 15 (6.1%) | |||
| Reduced | 105 | 14 (7.0–24.0) | 88 (84%) | 17 (16%) | 100 (95%) | 5 (4.8%) | |||
| Management duties (n = 1332) | 0.108 | 0.180 | 0.269 | ||||||
| None | 696 | 14 (6.0–24.0) | 584 (84%) | 112 (16%) | 641 (92%) | 55 (7.9%) | |||
| Some duties | 389 | 15 (7.0–26.0) | 309 (79%) | 80 (21%) | 367 (94%) | 22 (5.7%) | |||
| Main duty | 247 | 15 (8.0–25.5) | 203 (82%) | 44 (18%) | 233 (94%) | 14 (5.7%) | |||
| Work social support (n = 1332) | < 0.001 | < 0.001 | < 0.001 | ||||||
| High | 957 | 11 (6.0–20.0) | 861 (90%) | 96 (10%) | 911 (95%) | 46 (4.8%) | |||
| Medium | 353 | 25 (15.0–34.0) | 231 (65%) | 122 (35%) | 311 (88%) | 42 (12%) | |||
| Low | 22 | 36 (32.2–44.2) | 4 (18%) | 18 (82%) | 19 (86%) | 3 (14%) | |||
| Aggression from colleagues (n = 1332) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 1,214 | 13 (6.0–24.0) | 1,029 (85%) | 185 (15%) | 1,148 (95%) | 66 (5.4%) | |||
| Yes | 118 | 24.5 (15.2–37.0) | 67 (57%) | 51 (43%) | 93 (79%) | 25 (21%) | |||
| Fears losing job (n = 1332) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 1,173 | 13 (6.0–23.0) | 1,000 (85%) | 173 (15%) | 1,105 (94%) | 68 (5.8%) | |||
| Yes | 159 | 26 (16.0–34.5) | 96 (60%) | 63 (40%) | 136 (86%) | 23 (14%) | |||
| Feels overworked (n = 1332) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 857 | 10 (5.0–17.0) | 800 (93%) | 57 (6.7%) | 815 (95%) | 42 (4.9%) | |||
| Yes | 475 | 25 (16.0–34.0) | 296 (62%) | 179 (38%) | 426 (90%) | 49 (10%) | |||
| Works overtime (n = 1332) | < 0.001 | < 0.001 | 0.251 | ||||||
| Rarely/Never | 256 | 10.5 (5.0–19.0) | 230 (90%) | 26 (10%) | 241 (94%) | 15 (5.9%) | |||
| Occasionally | 499 | 12 (6.0–21.0) | 439 (88%) | 60 (12%) | 470 (94%) | 29 (5.8%) | |||
| Often | 577 | 19 (10.0–30.0) | 427 (74%) | 150 (26%) | 530 (92%) | 47 (8.1%) | |||
a Kruskal–Wallis rank sum test
bn (%), may not add up to 100% due to rounding
cPearson's Chi-squared test
Table 4.
Participant characteristics according to outcomes (sample source: Mobilized workers). Estimates are unweighted
| EE-MBI | Severe emotional exhaustion | Diagnosed burnout | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | N |
Score median (IQR) |
p-valuea | No N = 1000b |
Yes N = 245b |
p-valuec | No N = 1163b |
Yes N = 82b |
p-valuec |
| Teleworking group (n = 1245) | 0.009 | 0.102 | 0.093 | ||||||
| No change | 177 | 13 (8.0–23.0) | 152 (86%) | 25 (14%) | 168 (95%) | 9 (5.1%) | |||
| Decrease | 62 | 19.5 (12.0–29.5) | 46 (74%) | 16 (26%) | 53 (85%) | 9 (15%) | |||
| Increase | 261 | 16 (9.0–26.0) | 214 (82%) | 47 (18%) | 247 (95%) | 14 (5.4%) | |||
| Not possible | 572 | 17 (8.0–28.0) | 446 (78%) | 126 (22%) | 532 (93%) | 40 (7.0%) | |||
| Never | 173 | 12 (6.0–25.0) | 142 (82%) | 31 (18%) | 163 (94%) | 10 (5.8%) | |||
| Age group (n = 1245) | 0.006 | 0.512 | 0.563 | ||||||
| 25–34 | 107 | 17 (9.0–27.5) | 84 (79%) | 23 (21%) | 103 (96%) | 4 (3.7%) | |||
| 35–44 | 283 | 18 (9.0–28.0) | 222 (78%) | 61 (22%) | 266 (94%) | 17 (6.0%) | |||
| 45–54 | 489 | 16 (9.0–26.0) | 391 (80%) | 98 (20%) | 455 (93%) | 34 (7.0%) | |||
| 55–64 | 366 | 13 (7.0–24.0) | 303 (83%) | 63 (17%) | 339 (93%) | 27 (7.4%) | |||
| Sex (n = 1245) | < 0.001 | 0.110 | 0.014 | ||||||
| Male | 485 | 13 (7.0–25.0) | 401 (83%) | 84 (17%) | 464 (96%) | 21 (4.3%) | |||
| Female | 760 | 17 (9.0–27.0) | 599 (79%) | 161 (21%) | 699 (92%) | 61 (8.0%) | |||
| Education (n = 1245) | < 0.001 | < 0.001 | 0.191 | ||||||
| Tertiary | 799 | 16 (9.0–26.0) | 647 (81%) | 152 (19%) | 750 (94%) | 49 (6.1%) | |||
| Secondary | 419 | 15 (7.0–26.0) | 339 (81%) | 80 (19%) | 390 (93%) | 29 (6.9%) | |||
| Primary | 27 | 29 (18.0–39.5) | 14 (52%) | 13 (48%) | 23 (85%) | 4 (15%) | |||
| Household income (n = 1245) | 0.032 | 0.047 | 0.319 | ||||||
| High | 161 | 13 (7.0–24.0) | 138 (86%) | 23 (14%) | 152 (94%) | 9 (5.6%) | |||
| Middle | 711 | 15 (8.0–26.0) | 574 (81%) | 137 (19%) | 670 (94%) | 41 (5.8%) | |||
| Low | 179 | 18 (9.5–30.0) | 132 (74%) | 47 (26%) | 163 (91%) | 16 (8.9%) | |||
| Don't want to answer | 194 | 18 (9.0–27.0) | 156 (80%) | 38 (20%) | 178 (92%) | 16 (8.2%) | |||
| Financial security (n = 1245) | < 0.001 | < 0.001 | 0.139 | ||||||
| Can cover expenses | 580 | 16 (9.0–26.0) | 469 (81%) | 111 (19%) | 551 (95%) | 29 (5.0%) | |||
| Cautious / struggling | 157 | 23 (10.0–33.0) | 105 (67%) | 52 (33%) | 143 (91%) | 14 (8.9%) | |||
| Comfortable | 450 | 14 (8.0–24.0) | 381 (85%) | 69 (15%) | 417 (93%) | 33 (7.3%) | |||
| Don't want to answer | 58 | 17.5 (9.0–29.0) | 45 (78%) | 13 (22%) | 52 (90%) | 6 (10%) | |||
| Ethnicity (n = 1234) | 0.712 | 0.239 | 0.233 | ||||||
| European-Caucasian | 1,182 | 16 (8.0–26.0) | 954 (81%) | 228 (19%) | 1,107 (94%) | 75 (6.3%) | |||
| Other | 52 | 17 (9.0–30.0) | 38 (73%) | 14 (27%) | 46 (88%) | 6 (12%) | |||
| Living arrangement (n = 1245) | 0.189 | 0.107 | 0.563 | ||||||
| Couple without children | 279 | 15 (8.0–25.5) | 232 (83%) | 47 (17%) | 263 (94%) | 16 (5.7%) | |||
| Couple with children | 615 | 15 (8.0–26.0) | 503 (82%) | 112 (18%) | 578 (94%) | 37 (6.0%) | |||
| Single with children | 120 | 17 (8.8–29.0) | 92 (77%) | 28 (23%) | 109 (91%) | 11 (9.2%) | |||
| With other adults | 38 | 16 (9.2–29.5) | 28 (74%) | 10 (26%) | 36 (95%) | 2 (5.3%) | |||
| Alone | 193 | 17 (9.0–29.0) | 145 (75%) | 48 (25%) | 177 (92%) | 16 (8.3%) | |||
| Has young children (n = 1245) | 0.02 | 0.118 | 0.194 | ||||||
| No young children | 1,002 | 15 (8.0–26.0) | 814 (81%) | 188 (19%) | 931 (93%) | 71 (7.1%) | |||
| Has young children | 243 | 18 (9.0–29.0) | 186 (77%) | 57 (23%) | 232 (95%) | 11 (4.5%) | |||
| Overcrowded household (n = 1245) | 0.109 | 0.863 | 0.531 | ||||||
| Not overcrowded | 1,231 | 16 (8.0–26.0) | 988 (80%) | 243 (20%) | 1,151 (94%) | 80 (6.5%) | |||
| Overcrowded | 14 | 11.5 (8.2–15.2) | 12 (86%) | 2 (14%) | 12 (86%) | 2 (14%) | |||
| Noise level at home (n = 1245) | 0.085 | 0.081 | 0.012 | ||||||
| Not noisy | 425 | 15 (8.0–26.0) | 348 (82%) | 77 (18%) | 401 (94%) | 24 (5.6%) | |||
| A little noisy | 511 | 15 (9.0–26.0) | 412 (81%) | 99 (19%) | 477 (93%) | 34 (6.7%) | |||
| Moderately noisy | 258 | 16 (9.0–26.0) | 206 (80%) | 52 (20%) | 243 (94%) | 15 (5.8%) | |||
| Very noisy | 51 | 20 (8.0–33.0) | 34 (67%) | 17 (33%) | 42 (82%) | 9 (18%) | |||
| Access to a quiet room (n = 1245) | 0.024 | 0.046 | 0.447 | ||||||
| Yes | 1,004 | 15 (8.0–26.0) | 818 (81%) | 186 (19%) | 941 (94%) | 63 (6.3%) | |||
| No | 241 | 17 (9.0–29.0) | 182 (76%) | 59 (24%) | 222 (92%) | 19 (7.9%) | |||
| Self-reported general health (n = 1245) | < 0.001 | < 0.001 | 0.077 | ||||||
| Good | 1,160 | 15 (8.0–26.0) | 948 (82%) | 212 (18%) | 1,088 (94%) | 72 (6.2%) | |||
| Not good | 85 | 22 (12.0–36.0) | 52 (61%) | 33 (39%) | 75 (88%) | 10 (12%) | |||
| Self-reported mental state (n = 1245) | < 0.001 | < 0.001 | 0.060 | ||||||
| Good | 1,128 | 15 (8.0–25.0) | 929 (82%) | 199 (18%) | 1,059 (94%) | 69 (6.1%) | |||
| Not good | 117 | 25 (15.0–35.0) | 71 (61%) | 46 (39%) | 104 (89%) | 13 (11%) | |||
| Chronic condition (n = 1245) | 0.049 | 0.011 | 0.004 | ||||||
| No | 972 | 16 (8.0–26.0) | 796 (82%) | 176 (18%) | 919 (95%) | 53 (5.5%) | |||
| Yes | 273 | 16 (9.0–30.0) | 204 (75%) | 69 (25%) | 244 (89%) | 29 (11%) | |||
| Teleworking frequency (n = 1245) | 0.684 | 0.226 | 0.198 | ||||||
| Occasionally (irregular) | 256 | 16 (9.0–26.0) | 211 (82%) | 45 (18%) | 241 (94%) | 15 (5.9%) | |||
| 1 + days/week | 236 | 15.5 (8.0–25.0) | 196 (83%) | 40 (17%) | 221 (94%) | 15 (6.4%) | |||
| Every day | 8 | 19.5 (8.8–33.0) | 5 (63%) | 3 (38%) | 6 (75%) | 2 (25%) | |||
| Never | 745 | 16 (8.0–27.0) | 588 (79%) | 157 (21%) | 695 (93%) | 50 (6.7%) | |||
| Pandemic-related work change (n = 1245) | < 0.001 | < 0.001 | 0.788 | ||||||
| No change | 1,170 | 15 (8.0–26.0) | 952 (81%) | 218 (19%) | 1,094 (94%) | 76 (6.5%) | |||
| Change | 75 | 24 (14.0–33.5) | 48 (64%) | 27 (36%) | 69 (92%) | 6 (8.0%) | |||
| Change in contracted hours (n = 1245) | 0.003 | 0.031 | 0.045 | ||||||
| No change | 898 | 15 (8.0–26.0) | 734 (82%) | 164 (18%) | 831 (93%) | 67 (7.5%) | |||
| Increased | 255 | 18 (9.0–27.0) | 201 (79%) | 54 (21%) | 247 (97%) | 8 (3.1%) | |||
| Reduced | 92 | 20 (11.0–32.2) | 65 (71%) | 27 (29%) | 85 (92%) | 7 (7.6%) | |||
| Management duties (n = 1245) | 0.728 | 0.327 | 0.898 | ||||||
| None | 567 | 16 (8.0–26.0) | 464 (82%) | 103 (18%) | 529 (93%) | 38 (6.7%) | |||
| Some duties | 379 | 16 (9.0–26.0) | 304 (80%) | 75 (20%) | 353 (93%) | 26 (6.9%) | |||
| Main duty | 299 | 15 (8.0–27.0) | 232 (78%) | 67 (22%) | 281 (94%) | 18 (6.0%) | |||
| Work social support (n = 1245) | < 0.001 | < 0.001 | < 0.001 | ||||||
| High | 874 | 12 (7.0–22.0) | 766 (88%) | 108 (12%) | 834 (95%) | 40 (4.6%) | |||
| Medium | 346 | 24 (14.0–34.0) | 223 (64%) | 123 (36%) | 308 (89%) | 38 (11%) | |||
| Low | 25 | 31 (23.0–42.0) | 11 (44%) | 14 (56%) | 21 (84%) | 4 (16%) | |||
| Aggression from colleagues (n = 1245) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 1,116 | 15 (8.0–25.0) | 921 (83%) | 195 (17%) | 1,052 (94%) | 64 (5.7%) | |||
| Yes | 129 | 26 (16.0–34.0) | 79 (61%) | 50 (39%) | 111 (86%) | 18 (14%) | |||
| Fears losing job (n = 1245) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 1,111 | 14 (8.0–25.0) | 933 (84%) | 178 (16%) | 1,048 (94%) | 63 (5.7%) | |||
| Yes | 134 | 29.5 (18.2–36.8) | 67 (50%) | 67 (50%) | 115 (86%) | 19 (14%) | |||
| Feels overworked (n = 1245) | < 0.001 | < 0.001 | < 0.001 | ||||||
| No | 719 | 11 (5.0–19.0) | 664 (92%) | 55 (7.6%) | 687 (96%) | 32 (4.5%) | |||
| Yes | 526 | 24.5 (16.0–34.0) | 336 (64%) | 190 (36%) | 476 (90%) | 50 (9.5%) | |||
| Works overtime (n = 1245) | < 0.001 | < 0.001 | 0.698 | ||||||
| Rarely/Never | 246 | 13 (6.0–22.0) | 212 (86%) | 34 (14%) | 229 (93%) | 17 (6.9%) | |||
| Occasionally | 464 | 13 (7.0–23.0) | 401 (86%) | 63 (14%) | 437 (94%) | 27 (5.8%) | |||
| Often | 535 | 20 (11.0–30.0) | 387 (72%) | 148 (28%) | 497 (93%) | 38 (7.1%) | |||
aKruskal–Wallis rank sum test
bn (%), may not add up to 100% due to rounding
cPearson's Chi-squared test
The distribution of the exposure and outcomes also varied by occupation and job sector, for example most health professionals and personal care workers reported that it was not possible for them to telework (Table S3 to Table S6).
Regression analyses
When analyzing only the “population-based” participants (Fig. 2), and relative to those reporting no change in telework frequency relative to before the pandemic, a decrease in telework frequency was associated with higher EE-MBI scores (aβ 5.26 [95% CI: 1.47, 9.04]), as well as higher odds of diagnosed burnout (aOR 10.59 [3.24, 34.57]). An increase in telework frequency was not associated with EE-MBI scores, severe emotional exhaustion, or diagnosed burnout. Notably, those who reported that they “never” telework (but presumably had the option available) did not have any significantly different estimates for any of the outcomes. In contrast, and still relative to those reporting no change in telework frequency relative to before the pandemic, those reporting that it was “not possible” to telework had higher EE-MBI scores (aβ 3.51 [0.44, 6.59]), as well as higher odds of severe emotional exhaustion (aOR 2.10 [1.08, 4.08]) and diagnosed burnout (aOR 3.42 [1.22, 9.65]).
Fig. 2.
Weighted regression results for the population-based sample. Results of (a) the linear regression model with continuous EE-MBI score as the outcome, and the logistic regression models with (b) severe emotional exhaustion and (c) diagnosed burnout as the outcomes. Unadjusted estimates are from univariable models while adjusted estimates are from multivariable models that include all potential confounders as covariates: age, sex, education level, living arrangement, having young children, household density, noise levels at home, access to a quiet room, self-reported general health, occupation group, any pandemic-related work changes, and changes to contract percentages (further details in Table S1)
When analyzing only the “mobilized-workers” (Fig. 3), the estimates mirrored those of the “population-based” participants, with the exception that those who reported that teleworking was not possible did not have significantly different odds of diagnosed burnout relative to those who reported no change in telework frequency.
Fig. 3.
Unweighted regression results for the mobilized-workers sample. Results of (a) the linear regression model with continuous EE-MBI score as the outcome, and the logistic regression models with (b) severe emotional exhaustion and (c) diagnosed burnout as the outcomes. Unadjusted estimates are from univariable models while adjusted estimates are from multivariable models that include all potential confounders as covariates: age, sex, education level, living arrangement, having young children, household density, noise levels at home, access to a quiet room, self-reported general health, occupation group, any pandemic-related work changes, and changes to contract percentages (further details in Table S1)
Sensitivity analyses
In the sensitivity analyses keeping the “not possible” teleworking group as the reference category (Figure S3), those who report that they “never” telework (presumably by choice) had lower EE-MBI scores in both the “population-based” and “mobilized worker” samples, and lower odds of severe emotional exhaustion in the “population-based” sample only. After excluding those who reported either changes in their contract percentage (relative to baseline) or pandemic-related work changes, the direction and magnitude of the regression estimates largely echoed those of the main analyses, though with wider confidence intervals due to the reduced sample sizes (Figure S4).
Discussion
Key results
This study explored the relationship between changes in teleworking frequency, relative to before the COVID-19 pandemic, and two measures of occupational burnout: emotional exhaustion and diagnosed burnout. Using data from two distinct samples—population-based and mobilized workers—we found that compared with no change in teleworking frequency, a decrease in teleworking frequency was consistently associated with higher emotional exhaustion scores. In the population-based sample, but not in the mobilized-workers sample, those who indicated that telework was not possible also had higher odds of diagnosed burnout. In both samples, neither increases in teleworking frequency nor reporting “never” teleworking were significantly associated with differences in burnout outcomes compared to those reporting no change in teleworking frequency.
Interpretation
Our results align with existing literature on the impact of organizational change on employees mental health [33, 34], as changing telework arrangements can disrupt established work relationships, routines, and work-life boundaries [17, 18]. In the Job Demands-Resources framework, the psychological efforts on the part of the employee to adjust to these disruptions may act as organizational “demands” that exacerbate emotional exhaustion [34, 35].
Much of the pre-pandemic literature suggests that telework is largely well received and tends to mitigate emotional exhaustion [9, 19]. However, this mostly applies to voluntary telework, where employees for whom it is likely to be beneficial are more likely to opt for it and view it as an organizational resource. Pre-pandemic studies suggest that involuntary telework is associated with greater work-family conflict, stress, and burnout [36, 37]. Pandemic-related closures precipitated a largely mandatory shift to telework based on its technical feasibility rather than individual preferences or professional suitability [14, 38]. While this shift may not be associated with negative impacts on mental health early into the pandemic [39, 40], it’s important to recognize that individual circumstances likely influence the impacts on employee mental health [11].
In our study, participants reporting a decrease in telework likely consisted of workers who were already voluntarily teleworking before the pandemic, and so experiencing a reduction in telework may be more likely to be perceived as a reduction of an organizational “resource”. While our study did not explicitly explore voluntariness, compared with those who “never” telework we found that not having this possibility was associated with higher emotional exhaustion. One explanation is that while some in the latter group may wish to telework, they are unable to because of technical infeasibility or unwillingness on the part of the employer. Previous research suggests that regardless of whether or not they actually use the option, employees who have telework and other flexible working arrangements available report higher job satisfaction and organizational commitment, which are protective factors against emotional exhaustion and occupational burnout [16]. Future studies exploring the relationship between flexible working arrangements and employees’ wellbeing should explicitly account for the important dimension of voluntariness.
The benefits and drawbacks of telework are dependent on individual circumstances and personality traits [9, 11]. For example, teleworking imposes a certain degree of social separation from colleagues, and those living alone might be especially vulnerable to feelings of social isolation or a perceived lack of social support in the workplace [13]. For some, telework may facilitate work-life balance while for others it may blur the work-life boundary [41]. Living and family circumstances, for example caring for a small child at home or living in dense housing, can impact work-life conflict [41]. Some professions may be technically teleworkable, but if interpersonal interactions are an essential aspect of the role then teleworking can lead to a decline in quality and personal satisfaction [14]. As many organizations consider updating their telework policies in this “post-pandemic” period, they may benefit from engaging with employees to craft mutually beneficial arrangements [42, 43]. Future studies should consider including measures of personality traits to gain deeper insights into the individual circumstances that influence how people may benefit from telework and other flexible work arrangements.
Strengths and limitations
Strengths of the study include data collected from a large population-based study, which allowed adjustment for various sociodemographic, health, and work-related factors that were temporally measured before the outcomes. Importantly, the study sample includes economically active participants from a broad spectrum of occupations and job sectors. The study includes two samples, one that is population-based and for which estimates representative of the canton of Geneva could be obtained, and one that is based on a subset of workers who were likely mobilized in the early stages of the COVID-19 pandemic. The estimates from each sample are largely in agreement, providing additional confidence regarding the internal validity of the study. Another strength is the use of a validated scale to measure emotional exhaustion, a key component of occupational burnout. Professional diagnoses do not necessarily represent the true frequency of experienced burnout, for example due to variable help-seeking behavior and perceptions of stigma [44]. Instead, measurements of symptoms allow for routine screening to assess the work-environment and identify at-risk employees to provide help and prevent further progression of occupational burnout [4].
Our study has several limitations. First, questions on work-related emotional exhaustion and telework were posed only to respondents in salaried employment. This likely excluded people who left their job due to burnout, and this is supported in our data (Table S7), as respondents who met all the study criteria except for being in salaried employment were more likely to have been diagnosed with burnout within the past year. However, our primary focus was on current levels of work-related emotional exhaustion, making it relevant for actively employed participants. Second, our exposure and outcomes were measured in the same questionnaire, leaving open the possibility of reverse causality, for example respondents may have adjusted their telework habits after experiencing sustained emotional exhaustion or burnout. However, we anticipate negligible recall bias in measuring our exposure; given the pandemic’s sudden and significant societal impact, self-reported changes in telework frequency compared to pre-pandemic times should mostly reflect true changes. Third, the use of categorical data to measure changes in telework frequency (rather than, for example, querying the average number of hours of pre-pandemic and current telework) limits the precision and analytical possibilities of the findings. Fourth, participants who responded that they did not telework were not asked about changes in frequency relative to before the pandemic. This may result in some misclassification bias, as some of these participants are likely to have experienced a decrease in teleworking and now no longer telework by choice. The comparison between non-teleworking groups was not a primary focus, though we conducted sensitivity analyses to compare these groups, assuming that those who “never” telework likely have the option but choose not to. The results suggest that this group tends to report lower emotional exhaustion scores and lower odds of diagnosed burnout. Fifth, while we adjusted for several sociodemographic, family, health, and work-related factors, individual personality traits (for example stress-coping or openness to new experiences) were not measured in this study, though they are likely modulators of burnout [4]. Sixth, the self-reported nature of professionally diagnosed burnout may be subject to reporting errors or variability. While burnout is recognized in the International Classification of Diseases (ICD-11) [45] and the Systematized Nomenclature of Medicine Clinical Terms (SNOMED-CT), there is currently no universally accepted clinical diagnostic standard for burnout [1, 5, 46]. As a result, the criteria used by healthcare professionals to deliver a diagnosis may differ, leading to potential inconsistencies within our sample. This variability could affect the reliability of self-reported diagnoses and introduces the possibility of misclassification.
Conclusions
Our results suggest that in this post-pandemic context, a decrease in telework was associated with higher emotional exhaustion and odds of occupational burnout. Previous research suggests sustained emotional exhaustion can impact employee engagement, quality of work, and productivity. As many organizations consider work policy changes during this “post-pandemic” phase, especially those that would like a broader return to the office, it is important that they also consider the diverse psychological and emotional impacts of telework on employees.
Supplementary Information
Acknowledgements
The authors thank those who participated in the Specchio-COVID19 study, as well as all of those who made it possible, including the members of the Specchio-COVID19 study group: Isabelle Arm-Vernez, Andrew S. Azman, Delphine Bachmann, Antoine Bal, Jean-François Balavoine, Rémy P. Barbe, Hélène Baysson, Julie Berthelot, Gaëlle Bryand-Rumley, François Chappuis, Prune Collombet, Sophie Coudurier-Boeuf, Delphine S. Courvoisier, Carlos de Mestral, Paola D’ippolito, Richard Dubos, Roxane Dumont, Nacira El Merjani, Antoine Flahault, Natalie Francioli, Clément Graindorge, Idris Guessous, Séverine Harnal, Samia Hurst, Laurent Kaiser, Gabriel Kathari , Omar Kherad, Julien Lamour, Pierre Lescuyer, Arnaud G. L’Huillier, Andrea Jutta Loizeau, Elsa Lorthe, Chantal Martinez, Shannon Mechoullam, Ludovic Metral-Boffod, Mayssam Nehme, Natacha Noël, Francesco Pennacchio, Didier Pittet, Klara M. Posfay-Barbe, Géraldine Poulain, Caroline Pugin, Nick Pullen, Viviane Richard, Déborah Rochat, Khadija Samir, Hugo Santa Ramirez, Etienne Satin, Philippe Schaller, Stephanie Schrempft, Claire Semaani, Silvia Stringhini, Stéphanie Testini, Anshu Uppal, Déborah Urrutia-Rivas, Charlotte Verolet, Pauline Vetter, Jennifer Villers, Guillemette Violot, Nicolas Vuilleumier, Ania Wisniak, Sabine Yerly, and María-Eugenia Zaballa.
Declaration of Artificial Intelligence Generated Content (AIGC) tools in the writing process
During the preparation of this work, the authors used AI-assisted technologies to improve readability and language. After using these tools, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
Abbreviations
- aβ
Adjusted regression coefficients
- aOR
Adjusted odds ratios
- EE-MBI
Emotional exhaustion component of the Maslach Burnout Inventory.
- IQR
Interquartile range.
- ISCO-08
International Standard Classification of Occupations (2008).
Authors’ contributions
Conceptualization: A.U., N.P., I.G., S.S. and E.L.; Data curation: A.U., N.P. and J.L.; Formal analysis: A.U., N.P. and E.L.; Funding acquisition: I.G. and S.S.; Investigation: N.P., H.B., S.S. and M.Z.; Methodology: A.U., N.P. and E.L.; Project administration: H.B., M.N. and S.S.; Software: A.U. and N.P.; Supervision: N.P. and E.L.; Validation: N.P., H.B., S.S., A.R.B., M.Z., J.L., M.N., S.S. and E.L.; Visualization: A.U.; Writing – original draft: A.U.; Writing – review & editing: A.U., N.P., H.B., S.S., A.R.B., M.Z., J.L., M.N., I.G., S.S. and E.L. All authors read and approved the final manuscript.
Funding
Open access funding provided by University of Geneva The Specchio-COVID19 study was funded by the Swiss Federal Office of Public Health, the General Directorate of Health of the Department of Safety, Employment and Health of the canton of Geneva, the Private Foundation of the Geneva University Hospitals, the Swiss School of Public Health (Corona Immunitas Research Programme) and the Fondation des Grangettes. Award/Grant number are not applicable.
Data availability
The data that support the findings of this study are not openly available due to reasons of sensitivity and participant privacy, and are available from the corresponding author upon reasonable request.
Declarations
Ethics approval and consent to participate
This study was carried out in accordance with the Declaration of Helsinki. The Specchio-COVID19 study was approved by the Cantonal Research Ethics Commission of Geneva (CCER project ID 2020–00881). All participants provided electronic informed consent to participate in the digital cohort.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Elsa Lorthe, Email: elsa.lorthe@hug.ch.
Specchio-COVID19 study group:
Anshu Uppal, Nick Pullen, Hélène Baysson, Stephanie Schrempft, María-Eugenia Zaballa, Julien Lamour, Mayssam Nehme, Idris Guessous, Silvia Stringhini, Elsa Lorthe, Isabelle Arm-Vernez, Andrew S. Azman, Delphine Bachmann, Antoine Bal, Jean-François Balavoine, Rémy P. Barbe, Julie Berthelot, Gaëlle Bryand-Rumley, François Chappuis, Prune Collombet, Sophie Coudurier-Boeuf, Delphine S. Courvoisier, Carlos de Mestral, Paola D’ippolito, Richard Dubos, Roxane Dumont, Nacira El Merjani, Antoine Flahault, Natalie Francioli, Clément Graindorge, Séverine Harnal, Samia Hurst, Laurent Kaiser, Gabriel Kathari, Omar Kherad, Pierre Lescuyer, Arnaud G. L’Huillier, Andrea Jutta Loizeau, Chantal Martinez, Shannon Mechoullam, Ludovic Metral-Boffod, Natacha Noël, Francesco Pennacchio, Didier Pittet, Klara M. Posfay-Barbe, Géraldine Poulain, Caroline Pugin, Viviane Richard, Déborah Rochat, Khadija Samir, Hugo Santa Ramirez, Etienne Satin, Philippe Schaller, Claire Semaani, Stéphanie Testini, Déborah Urrutia-Rivas, Charlotte Verolet, Pauline Vetter, Jennifer Villers, Guillemette Violot, Nicolas Vuilleumier, Ania Wisniak, and Sabine Yerly
References
- 1.Guseva Canu I, Marca SC, Dell’Oro F, Balázs Á, Bergamaschi E, Besse C, et al. Harmonized definition of occupational burnout: A systematic review, semantic analysis, and Delphi consensus in 29 countries. Scand J Work Environ Health. 2021;47(2):95–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ahola K, Honkonen T, Isometsä E, Kalimo R, Nykyri E, Koskinen S, et al. Burnout in the general population. Soc Psychiatry Psychiatr Epidemiol. 2006;41(1):11–7. [DOI] [PubMed] [Google Scholar]
- 3.Salvagioni DAJ, Melanda FN, Mesas AE, González AD, Gabani FL, de Andrade SM. Physical, psychological and occupational consequences of job burnout: A systematic review of prospective studies. PLoS ONE. 2017;12(10): e0185781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Edú-Valsania S, Laguía A, Moriano JA. Burnout: A Review of Theory and Measurement. Int J Environ Res Public Health. 2022;19(3):1780. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Al-Gobari M, Shoman Y, Blanc S, Canu IG. Point prevalence of burnout in Switzerland: a systematic review and meta-analysis. Swiss Med Wkly. 2022;152: w30229. [DOI] [PubMed] [Google Scholar]
- 6.Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The job demands-resources model of burnout. J Appl Psychol. 2001;86(3):499–512. [PubMed] [Google Scholar]
- 7.Bakker AB, Demerouti E, Euwema MC. Job resources buffer the impact of job demands on burnout. J Occup Health Psychol. 2005;10(2):170–80. [DOI] [PubMed] [Google Scholar]
- 8.Eurofound. The rise in telework: impact on working conditions and regulations. [Internet]. Luxembourg: Publications Office of the European Union; 2022. Available from: https://data.europa.eu/doi/10.2806/069206. Cited 2023 Jun 27.
- 9.Allen TD, Golden TD, Shockley KM. How Effective Is Telecommuting? Assessing the Status of Our Scientific Findings. Psychol Sci Public Interest. 2015;16(2):40–68. [DOI] [PubMed] [Google Scholar]
- 10.Wöhner F. Work flexibly, travel less? The impact of telework and flextime on mobility behavior in Switzerland. J Transp Geogr. 2022;102: 103390. [Google Scholar]
- 11.Lunde LK, Fløvik L, Christensen JO, Johannessen HA, Finne LB, Jørgensen IL, et al. The relationship between telework from home and employee health: a systematic review. BMC Public Health. 2022;22(1):47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Golden TD. Avoiding depletion in virtual work: Telework and the intervening impact of work exhaustion on commitment and turnover intentions. J Vocat Behav. 2006;69(1):176–87. [Google Scholar]
- 13.Vander Elst T, Verhoogen R, Sercu M, Van den Broeck A, Baillien E, Godderis L. Not Extent of Telecommuting, But Job Characteristics as Proximal Predictors of Work-Related Well-Being. J Occup Environ Med. 2017;59(10): e180. [DOI] [PubMed] [Google Scholar]
- 14.Sostero M, Milasi S, Hurley J, Fernandez-Macías E, Bisello M. Teleworkability and the COVID-19 crisis: a new digital divide? [Internet]. European Commission; 2020. Report No.: JRC121193. Available from: https://joint-research-centre.ec.europa.eu/publications/teleworkability-and-covid-19-crisis-new-digital-divide_en
- 15.Gibson CB, Gilson LL, Griffith TL, O’Neill TA. Should employees be required to return to the office? Organ Dyn. 2023;52(2): 100981. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Chen Y, Fulmer IS. Fine-tuning what we know about employees’ experience with flexible work arrangements and their job attitudes. Hum Resour Manage. 2018;57(1):381–95. [Google Scholar]
- 17.Fan W, Moen P. Ongoing Remote Work, Returning to Working at Work, or in between during COVID-19: What Promotes Subjective Well-Being? J Health Soc Behav. 2023Mar;64(1):152–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Pandita D, Gupta D, Vapiwala F. Rewinding Back into the Old Normal: Why is Return-to-Office Stressing Employees Out? Empl Responsib Rights J. 2024. Available from: https://link.springer.com/10.1007/s10672-023-09491-0. Cited 2024 Nov 21.
- 19.Sardeshmukh SR, Sharma D, Golden TD. Impact of telework on exhaustion and job engagement: a job demands and job resources model. New Technol Work Employ. 2012;27(3):193–207. [Google Scholar]
- 20.Baysson H, Pennachio F, Wisniak A, Zabella ME, Pullen N, Collombet P, et al. Specchio-COVID19 cohort study: a longitudinal follow-up of SARS-CoV-2 serosurvey participants in the canton of Geneva, Switzerland. BMJ Open. 2022;12(1): e055515. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Stringhini S, Wisniak A, Piumatti G, Azman AS, Lauer SA, Baysson H, et al. Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Geneva, Switzerland (SEROCoV-POP): a population-based study. The Lancet. 2020;396(10247):313–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Stringhini S, Zaballa ME, Perez-Saez J, Pullen N, de Mestral C, Picazio A, et al. Seroprevalence of anti-SARS-CoV-2 antibodies after the second pandemic peak. Lancet Infect Dis. 2021;21(5):600–1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Stringhini S, Zaballa ME, Pullen N, Perez-Saez J, de Mestral C, Loizeau AJ, et al. Seroprevalence of anti-SARS-CoV-2 antibodies 6 months into the vaccination campaign in Geneva, Switzerland, 1 June to 7 July 2021. Eurosurveillance. 2021;26(43):2100830. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zaballa ME, Perez-Saez J, de Mestral C, Pullen N, Lamour J, Turelli P, et al. Seroprevalence of anti-SARS-CoV-2 antibodies and cross-variant neutralization capacity after the Omicron BA.2 wave in Geneva, Switzerland: a population-based study. Lancet Reg Health Eur. 2023;24:100547. [DOI] [PMC free article] [PubMed]
- 25.Stringhini S, Zaballa ME, Pullen N, de Mestral C, Perez-Saez J, Dumont R, et al. Large variation in anti-SARS-CoV-2 antibody prevalence among essential workers in Geneva, Switzerland. Nat Commun. 2021;12(1):3455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Maslach C, Jackson SE, Leiter MP. Maslach Burnout Inventory Manual. 4th ed. Menlo Park, CA: Mind Garden Inc; 2016. [Google Scholar]
- 27.Dion G, Tessier R. Validation de la traduction de l’Inventaire d’épuisement professionnel de Maslach et Jackson [Validation of a French translation of the Maslach Burnout Inventory (MBI).]. Can J Behav Sci Rev Can Sci Comport. 1994;26:210–27. [Google Scholar]
- 28.International Labour Organization. International Standard Classification of Occupations. 2008. Available from: https://www.ilo.org/public/english/bureau/stat/isco/isco08/. Cited 2023 Jul 3.
- 29.Office Cantonal des Statistiques (OCSTAT). T 20.02.7.01 - Quantiles du revenu annuel brut des ménages selon le type de ménage, en 2015–2017. Canton de Genève; 2020. Available from: https://statistique.ge.ch/domaines/20/20_02/tableaux.asp. Cited 2023 Jun 13.
- 30.Karasek R, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B. The Job Content Questionnaire (JCQ): An instrument for internationally comparative assessments of psychosocial job characteristics. J Occup Health Psychol. 1998;3(4):322–55. [DOI] [PubMed] [Google Scholar]
- 31.Bué J, Coutrot T, Guignon N, Sandret N. Les facteurs de risques psychosociaux au travail: Une approche quantitative par l’enquête Sumer. Rev Fr Aff Soc. 2008Jun 1;2:45–70. [Google Scholar]
- 32.Westreich D, Greenland S. The Table 2 Fallacy: Presenting and Interpreting Confounder and Modifier Coefficients. Am J Epidemiol. 2013;177(4):292–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bamberger SG, Vinding AL, Larsen A, Nielsen P, Fonager K, Nielsen RN, et al. Impact of organisational change on mental health: a systematic review. Occup Environ Med. 2012;69(8):592–8. [DOI] [PubMed] [Google Scholar]
- 34.Day A, Crown SN, Ivany M. Organisational change and employee burnout: The moderating effects of support and job control. Saf Sci. 2017;100:4–12. [Google Scholar]
- 35.Hetty van Emmerik IJ, Bakker AB, Euwema MC. Explaining employees’ evaluations of organizational change with the job‐demands resources model. Career Dev Int. 2009;14(6):594–613.
- 36.Lapierre LM, van Steenbergen EF, Peeters MCW, Kluwer ES. Juggling work and family responsibilities when involuntarily working more from home: A multiwave study of financial sales professionals. J Organ Behav. 2016;37(6):804–22. [Google Scholar]
- 37.Kaduk A, Genadek K, Kelly EL, Moen P. Involuntary vs. Voluntary Flexible Work: Insights for Scholars and Stakeholders. Community Work Fam. 2019;22(4):412–42. [DOI] [PMC free article] [PubMed]
- 38.Kaluza AJ, van Dick R. Telework at times of a pandemic: The role of voluntariness in the perception of disadvantages of telework. Curr Psychol. 2022;42:18578–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.McDowell CP, Herring MP, Lansing J, Brower CS, Meyer JD. Associations Between Employment Changes and Mental Health: US Data From During the COVID-19 Pandemic. Front Psychol. 2021;12: 631510. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Schrempft S, Pullen N, Baysson H, Wisniak A, Zaballa ME, Pennacchio F, et al. Prevalence and predictors of psychological distress before, during, and after a COVID-19 pandemic wave in Switzerland, 2021. J Psychiatr Res. 2023;158:192–201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Elbaz S, Richards JB, Provost SY. Teleworking and work–life balance during the COVID-19 pandemic: A scoping review. Can Psychol Psychol Can. 2023;64(4):227–58. [Google Scholar]
- 42.Bordia P, Hobman E, Jones E, Gallois C, Callan VJ. Uncertainty During Organizational Change: Types, Consequences, and Management Strategies. J Bus Psychol. 2004;18(4):507–32. [Google Scholar]
- 43.Li JY, Sun R, Tao W, Lee Y. Employee coping with organizational change in the face of a pandemic: The role of transparent internal communication. Public Relat Rev. 2021;47(1): 101984. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Dyrbye LN, Leep Hunderfund AN, Winters RC, Moeschler SM, Vaa Stelling BE, Dozois EJ, et al. The Relationship Between Burnout and Help-Seeking Behaviors, Concerns, and Attitudes of Residents. Acad Med J Assoc Am Med Coll. 2021;96(5):701–8. [DOI] [PubMed] [Google Scholar]
- 45.WHO. International Classification of Diseases, Eleventh Revision (ICD-11). 2019. Available from: https://icd.who.int/browse11. Cited 2024 Nov 26.
- 46.Heinemann LV, Heinemann T. Burnout Research: Emergence and Scientific Investigation of a Contested Diagnosis. SAGE Open. 2017Jan;7(1):2158244017697154. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are not openly available due to reasons of sensitivity and participant privacy, and are available from the corresponding author upon reasonable request.