Modelling self-diagnosed burnout as a categorical syndrome

• Published in: Acta neuropsychiatrica 2023-02, Vol.35 (1), p.50-58
• Main Authors: Tavella, Gabriela, Spoelma, Michael J, Hadzi-Pavlovic, Dusan, Bayes, Adam, Jebejian, Artin, Manicavasagar, Vijaya, Walker, Peter, Parker, Gordon
• Format: Article
• Language: English
• Subjects: Adjustment; Burnout; Burnout, Professional - diagnosis; Burnout, Professional - epidemiology; Burnout, Professional - psychology; Humans; Mental depression; Mental Disorders; Mental health; Pandemics; Surveys and Questionnaires
• ISSN: 0924-2708; 1601-5215
• DOI: 10.1017/neu.2022.25

There is currently little consensus as to how burnout is best defined and measured, and whether the syndrome should be afforded clinical status. The latter issue would be advanced by determining whether burnout is a singular dimensional construct varying only by severity (and with some level of severity perhaps indicating clinical status), or whether a categorical model is superior, presumably reflecting differing 'sub-clinical' versus 'clinical' or 'burning out' vs 'burnt out' sub-groups. This study sought to determine whether self-diagnosed burnout was best modelled dimensionally or categorically. We recently developed a new measure of burnout which includes symptoms of exhaustion, cognitive impairment, social withdrawal, insularity, and other psychological symptoms. Mixture modelling was utilised to determine if scores from 622 participants on the measure were best modelled dimensionally or categorically. A categorical model was supported, with the suggestion of a sub-syndromal class and, after excluding such putative members of that class, two other classes. Analyses indicated that the latter bimodal pattern was not likely related to current working status or differences in depression symptomatology between participants, but reflected subsets of participants with and without a previous diagnosis of a mental health condition. Findings indicated that sub-categories of self-identified burnout experienced by the lay population may exist. A previous diagnosis of a mental illness from a mental health professional, and therefore potentially a psychological vulnerability factor, was the most likely determinant of the bimodal data, a finding which has theoretical implications relating to how best to model burnout.