Effectiveness of an online recovery training for employees exposed to blurred boundaries between work and non-work: Bayesian analysis of a randomised controlled trial

Hanna Brückner; Sebastian Wallot; Hanne Horvath; David Daniel Ebert; Dirk Lehr

doi:10.1136/bmjment-2024-301016

Effectiveness of an online recovery training for employees exposed to blurred boundaries between work and non-work: Bayesian analysis of a randomised controlled trial

BMJ Ment Health. 2024 Apr 19;27(1):e301016. doi: 10.1136/bmjment-2024-301016.

Authors

Hanna Brückner¹, Sebastian Wallot², Hanne Horvath³, David Daniel Ebert⁴, Dirk Lehr⁵

Affiliations

¹ Institute for Sustainability Education and Psychology, Department of Health Psychology and Applied Biological Psychology, Leuphana University, Lüneburg, Germany hanna.brueckner@leuphana.de.
² Institute for Sustainability Education and Psychology, Department of Methodology and Evaluation Research, Leuphana University, Lüneburg, Germany.
³ GET.ON Institute for Online Health Trainings GmbH, Berlin, Germany.
⁴ Institute for Psychology & Digital Mental Health Care, Department of Sports and Health Sciences, Technical University, Munich, Germany.
⁵ Institute for Sustainability Education and Psychology, Department of Health Psychology and Applied Biological Psychology, Leuphana University, Lüneburg, Germany.

Abstract

Background: Blurred work-non-work boundaries can have negative effects on mental health, including sleep.

Objectives: In a randomised control trial, we aimed to assess the effectiveness of an online recovery training programme designed to improve symptoms of insomnia in a working population exposed to blurred boundaries.

Methods: 128 participants with severe insomnia symptoms (Insomnia Severity Index ≥15) and working under blurred work and non-work conditions (segmentation supplies <2.25) were randomly assigned to either the recovery intervention or a waitlist control group (WLC). The primary outcome was insomnia severity, assessed at baseline, after 2 months (T2) and 6 months (T3).

Findings: A greater reduction in insomnia was observed in the intervention compared with the WLC group at both T2 (d=1.51; 95% CI=1.12 o 1.91) and T3 (d=1.63; 95% CI=1.23 to 2.03]. This was shown by Bayesian analysis of covariance (ANCOVA), whereby the ANCOVA model yielded the highest Bayes factor (BF ₁₀=3.23×e⁶⁰] and a 99.99% probability. Likewise, frequentist analysis revealed significantly reduced insomnia at both T2 and T3. Beneficial effects were found for secondary outcomes including depression, work-related rumination, and mental detachment from work. Study attrition was 16% at T2 and 44% at T3.

Conclusions: The recovery training was effective in reducing insomnia symptoms, work related and general indicators of mental health in employees exposed to blurred boundaries, both at T2 and T3.

Clinical implications: In addition to demonstrating the intervention's effectiveness, this study exemplifies the utilisation of the Bayesian approach in a clinical context and shows its potential to empower recipients of interventional research by offering insights into result probabilities, enabling them to draw informed conclusions.

Trial registration number: German Clinical Trial Registration (DRKS): DRKS00006223, https://drks.de/search/de/trial/DRKS00006223.

Keywords: Data Interpretation, Statistical; Depression & mood disorders; Sleep.

Publication types

Randomized Controlled Trial

MeSH terms

Bayes Theorem
Cognitive Behavioral Therapy*
Data Collection
Humans
Sleep
Sleep Initiation and Maintenance Disorders* / therapy