To the Editor:

Strangely enough, diagnostic criteria for burnout failed to be established in more than 40 years of sustained research on the syndrome [1]. Arguably an even stranger phenomenon is the widespread tendency among burnout researchers to put the cart before the horse by trying to estimate the prevalence of a syndrome that cannot be formally diagnosed [1, 2]. Such attempts typically involve the use of arbitrary categorization criteria, cobbled together without any clear rationale. In the present letter, we discuss a recently published study by Holmes et al. [3] that, in our estimation, is symptomatic of this confusing trend.

Holmes et al. [3] addressed the issue of burnout among residents of various medical specialties. Burnout was assessed with the emotional exhaustion and depersonalization subscales of the Maslach Burnout Inventory (MBI), a self-administered questionnaire [4]. The authors found that 69 % of the surveyed residents suffered from burnout. Burnout rates ranged from 46 to 89 % among the medical specialties of interest. The authors additionally reported that directors of residency programs underestimated the prevalence of burnout among residents. From the authors’ viewpoint, the obtained results “clearly demonstrate that resident burnout is a highly prevalent problem.” We do not share such conclusions.

In order to identify “residents with burnout,” Holmes et al. [3] used cut-off values available in the manual of the MBI [4]. Unfortunately, these cut-off values have been expressly indicated to be unsuited for diagnostic purposes (p. 9) [4]. Indeed, the cut-points in question reflect a mere tercile-based split and have no clinical or theoretical underpinning [2]. Furthermore, the authors allowed the respondents to be categorized as “burned out” based on their score on either the emotional exhaustion or the depersonalization subscale of the MBI. By doing so, the authors treated emotional exhaustion and depersonalization as two independent entities, rather than as the components of a unified syndrome—burnout. If, in the authors’ mind, emotional exhaustion and depersonalization represent separate entities that do not need to be studied in combination, the reason for grouping them under a single label, “burnout,” is arcane.

The authors justified the use of their categorization criteria by a will to be “consistent with the literature.” However, the use of similar categorization criteria in past research does not change anything to their problematic character. The methods employed in past studies should be critically analyzed, rather than merely reused, if burnout research is to progress. Some investigators may argue that the use of the same categorization criteria from one study to another at least promotes between-study comparability. We do not think that comparing arbitrary estimates with other arbitrary estimates is a satisfactory investigation method.

In view of the aforementioned problems, the authors’ observation that residency program directors underestimated the prevalence of burnout among residents loses its meaning. Indeed, for such an observation to be relevant, a valid comparison point is needed. As previously underscored, no such point was provided. The very idea of questioning residency program directors about the “rates of burnout” in residents is perplexing given that, most probably, all directors do not have the same understanding of what “burnout” means.

Burnout has become a fashionable construct in occupational health research. The characterization of the burnout syndrome, however, remains worryingly deficient [5]. Instead of multiplying studies of burnout’s “prevalence,” burnout investigators should concentrate their efforts on clarifying the nosological status of the entity that they purport to examine. Interestingly, Holmes et al. [3] urged psychiatrists to “provide guidance to the health care system on how to address the complicated problem of physician burnout.” In order to tackle the question of physician burnout, psychiatrists would first need to know what constitutes a case of burnout and how such a case can be (differentially) diagnosed.