As expected, this study reported a high rate of depressive symptoms among intensivists. The presence of depressive symptoms was not associated with patients’ severity of illness. In contrast, depression in intensivists was associated with organizational factors, such as workload and relationships with colleagues.
Approximately a quarter of the physicians included in the present study exhibited signs of depression. Coomber et al.
 reported that 12% of U.K. ICU physicians showed clinically important levels of depression. In a 2002 survey of internal medicine trainees, 40% of female residents and 32% of male residents reported four or five symptoms of depression
. It is very interesting to observe that the respondents with depressive symptoms included in the present study had a longer ICU practice compared with those who had not.
In the general population, risk factors for major depression include being a woman
. Frank and Dingle reported that women physicians with histories of self-identified depression shared many attributes, such as not being partnered, or being dissatisfied with career and work
. This was not the case in the present study where gender was not associated with depressive symptoms.
In the study by Coomber et al.
, there was no relationship between the level of depression and the age, the reported hours worked in the previous week, or the number of ICU beds for which they were responsible. This lack of association with work hours has been noted so far
. In the present study, we did not report any association between work hours and depression. However, there was an inverse relationship between symptoms of depression and the delay since the last nonworking weekend, suggesting that a sufficient period of rest is probably a more important factor than the number of work hours. ICU organizational aspects therefore are very important and could be a major issue in decreasing physicians’ depression. Chronic sleep deprivation may be a contributive factor that explains such a high rate of physicians presenting depressive symptoms
[16–18]. Longer resting periods could be therefore useful to improve physicians’ psychological health. Depressive disorders are likely associated to job-related burnout. More specifically some studies suggest that burnout leads to depressive symptoms
[19, 20]. In the present study, the presence of a high level of burnout was strongly related to the presence of depressive symptoms.
The limitations of this study include the biases of self-reporting (e.g., skipping questions, nondisclosure, same day of completion of the questionnaire by the intensivists and different interpretations of meaning), the lack of standardized interviews or specific criteria for making psychiatric diagnoses, incomplete information (e.g., no items on mental health treatment). The responders may not represent the whole French ICU physician community in terms of psychiatric morbidity, but we had no way of exploring the state of mental health of the nonresponders. Nonetheless, we would argue that the level of morbidity we have detected is likely to be a conservative estimate, because it can be argued that nonresponse is associated with “burnout” and “depression.” However, we do recognize that such conjecture needs further investigation, because nonresponse may be simple disinterest rather than overwork and lack of time. The study was done in 2004. The working conditions may have change since 2004, in particular concerning the number of ICU physicians who leave the hospital after a night shift. This could have modified the prevalence of depressive symptoms. However, we have no available information to indicate that the rate of intensivists leaving the hospital early in the morning following a night shift has increased recently. Additional investigations are necessary to design appropriate interventions that could be implemented to decrease depression rate in ICU doctors.
It is important to diagnose and to treat depression, because treatment is associated with improved work productivity
, and with a reduction of suicide
. Interventional studies designed to evaluate whether alteration of ICU organization is able to decrease the prevalence/incidence of depressive symptoms are warranted. It is an important objective individually (to increase the intensivist well-being) but also collectively (ICU caregivers and patients) by potentially reducing intensivists’ turnover.