The Occupational Depression Inventory—a solution for estimating the prevalence of job-related distress

Yang et al. (2021) examined the mental health status of community epidemic prevention workers during the postpandemic era of COVID-19 in a city in China. The authors attempted to estimate the prevalence of several mental health conditions, including burnout—a purported indicator of job-related distress. Yang et al. (2021) used the Maslach Burnout Inventory-General Survey (MBI-GS) to identify “cases” of burnout and concluded that the prevalence of burnout exceeded 50% in their sample. We applaud Yang et al.’s (2021) effort to provide information on the mental health status of community epidemic prevention workers within the postpandemic context. We are concerned, however, about the authors’ reliance on the burnout construct and the MBI-GS to estimate the prevalence of job-related distress. In this paper, we first describe some of the problems plaguing the burnout construct and its measures. Then, we present the Occupational Depression Inventory, a new instrument designed to help occupational health specialists get a clearer view of the mental health status of the workforce (Bianchi and Schonfeld, 2020).

As underscored by many investigators over the years (e.g., Bianchi et al., 2021), the very idea of estimating the prevalence of burnout is questionable given the absence of established diagnostic criteria for the syndrome. The MBI(-GS) was not designed for diagnostic purposes and alternative measures of burnout do not make the problem of burnout's diagnosis more tractable. Research on burnout's prevalence has been strongly criticized for its use of clinically and theoretically arbitrary identification criteria (e.g., tercile-based splits) leading to virtually uninterpretable results (Rotenstein et al., 2018). It is of note that the identification criteria used in burnout research have not only been marked by arbitrariness; they have also involved considerable heterogeneity, hindering between-study comparisons. As an illustration, Rotenstein et al. (2018) identified no fewer than 142 unique characterizations of burnout in a systematic review of 182 studies dedicated to the prevalence of burnout among physicians. Unsurprisingly, the estimates produced vary dramatically as a function of how burnout is defined—e.g., from 3% to 91% in a large-sample study by Hewitt et al. (2020). Such a state of affairs is disquieting and ultimately prevents occupational health specialists from making informed and authoritative decisions regarding the problem of job-related distress. We stress that the impossibility of diagnosing burnout has ramifications that go far beyond the prevalence issue. For example, the non-diagnosability of burnout impedes investigators’ ability to identify biological markers for the syndrome or to develop effective treatments and interventions.

The confusion surrounding the characterization of burnout is probably unsurprising given the origin of the entity. The burnout construct and the MBI were not rooted in any particular theory or grounded in thorough clinical investigations. The burnout construct was introduced in the literature based on anecdotal evidence, with no reference to the already well-developed research on stress-related conditions (e.g., research on learned helplessness). The originality of burnout was taken for granted rather than demonstrated. There is now substantial evidence that what pioneers of burnout research approached as a new and unique phenomenon is best understood as a depressive response to unresolvable (job) stress (e.g., Bianchi et al., 2017, 2021).

Burnout's conceptualization and measurement have been deeply problematic. It is in this context that we recently developed the Occupational Depression Inventory (Bianchi and Schonfeld, 2020). The Occupational Depression Inventory is a dual-purpose instrument that allows investigators to (a) grade the severity of work-attributed depressive symptoms (dimensional approach) and (b) identify likely cases of occupational depression with reference to internationally recognized diagnostic criteria for major depressive disorder (categorical approach). Anchored in the well-established area of stress and depression research, the Occupational Depression Inventory benefits from solid clinical and theoretical foundations. Available evidence indicates that the Occupational Depression Inventory has excellent psychometric and structural properties—arguably much stronger than those of the MBI. Importantly, by referencing the nine main symptoms of major depression, the Occupational Depression Inventory has a broader symptom coverage compared to burnout scales. For instance, the Occupational Depression Inventory assesses work-attributed suicidal thoughts, a critical sign of job-related distress having potentially lethal consequences. Burnout measures overlook such crucial symptoms. By contrast with measures such as the MBI, the Occupational Depression Inventory can be used at no cost, which is nonnegligible considering that researchers’ resources are limited.

Estimating the prevalence of job-related distress has been challenging for occupational health specialists, including psychiatrists specialized in the domain. The Occupational Depression Inventory constitutes a promising solution to the problem of estimating the prevalence of job-related distress. We do not believe that continuing to rely on burnout will be helpful in light of the multiple flaws and shortcomings of the construct. A change in paradigm from burnout to occupational depression may considerably strengthen our ability to help stressed-out workers and combat depressogenic working conditions (e.g., depressogenic management styles). It should be kept in mind that occupational health specialists do not operate in a vacuum. Their action is embedded in a complex web of hierarchical relationships. In such a context, the importance of reliable and valid assessment devices cannot be overstated if occupational health specialists’ recommendations for organizational changes are to be more than mere incantations.