Abstract
Most research on burnout is based on self-reported questionnaires. Nevertheless, as far as the clinical judgement is concerned, a lack of consensus about burnout diagnosis constitutes a risk of misdiagnosis. Hence, this study aims to assess the added value of a joint use of two tools and compare their diagnostic accuracy: (1) the early detection tool of burnout, a structured interview guide, and (2) the Oldenburg burnout inventory, a self-reported questionnaire. The interview guide was tested in 2019 by general practitioners and occupational physicians among 123 Belgian patients, who also completed the self-reported questionnaire. A receiver operating characteristic curve analysis allowed the identification of a cut-off score for the self-reported questionnaire. Diagnostic accuracy was then contrasted by a McNemar chi-squared test. The interview guide has a significantly higher sensitivity (0.76) than the self-reported questionnaire (0.70), even by comparing the self-reported questionnaires with the interviews of general practitioners and occupational physicians separately. However, both tools have a similar specificity (respectively, 0.60-0.67), except for the occupational physicians' interviews, where the specificity (0.68) was significantly lower than the self-reported questionnaire (0.70). In conclusion, the early detection tool of burnout is more sensitive than the Oldenburg burnout inventory, but seems less specific. However, by crossing diagnoses reported by patients and by physicians, they both seem useful to support burnout diagnosis.