Imo ^{Reference Imo1} conducted a systematic literature review of research on the prevalence of burnout among UK medical doctors, arriving at the conclusion that the prevalence of burnout in this population is ‘worryingly high’. Problematically, it turns out that such a conclusion cannot be drawn in view of the state of burnout research. Indeed, there are no clinically valid, commonly shared diagnostic criteria for burnout. ^{Reference Bianchi, Schonfeld and Laurent2,Reference Bianchi, Schonfeld and Laurent3} Given that what constitutes a case of burnout is undefined, how could an investigator estimate the prevalence of burnout, let alone conclude that burnout is widespread? As demonstrated elsewhere, ^{Reference Bianchi, Schonfeld and Laurent2–Reference Bianchi, Schonfeld and Laurent5} the diffuse estimates of burnout prevalence actually rely on categorisation criteria that are nosologically arbitrary and devoid of any sound theoretical justification. It is disconcerting to observe that studies of burnout prevalence continue multiplying in spite of the publication of several warnings against such research practices ^{Reference Bianchi, Schonfeld and Laurent2–Reference Weber and Jaekel-Reinhard6}

Another problem bearing on Imo's conclusions ^{Reference Imo1} lies in the unknown representativeness (e.g. in terms of gender, age, place of residence, or family status) of the samples of UK medical doctors surveyed in burnout research. Although the author partly acknowledges this problem in the limitation section of his article, he does not seem to take full account of the consequences of such a state of affairs. This state of affairs implies that the results of the reviewed studies cannot be generalised to the population of UK medical doctors.

All in all, the review ^{Reference Imo1} is undermined by the very research it relies on. We recommend that researchers interested in burnout start at the beginning, that is to say, by establishing a reasoned, clinically founded (differential) diagnosis for their entity of interest. As long as investigators do not complete the required groundwork for establishing a diagnosis and remain unable to distinguish a case of burnout from either a non-case or an existing disorder, conclusions regarding the prevalence of burnout will be nonsense. An immediately available solution for effectively monitoring and protecting physicians' occupational health would be to shift our focus from burnout to job-related depression. ^{Reference Bianchi, Schonfeld and Laurent2,Reference Bianchi, Schonfeld, Vandel and Laurent7}