Covid-19 pandemic-related changes in teleworking, emotional exhaustion, and occupational burnout: a cross-sectional analysis of a cohort study

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.

Keywords: Diagnosed burnout; Emotional exhaustion; Organizational change; Remote work; Work from home.

Fig. 1

Flowchart of occupational health questionnaire invitations and inclusion into analytical sample

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)

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)