Anxiety is commonly considered as an epiphenomenon of the cognitive and emotional response to a threat. Anxiety is a consequence of the reciprocal interaction between stress and the neuroendocrine, autonomic and immune systems. A systematic and circumstantial assessment of anxiety in critically ill patients has never been done. Our hypothesis is that high levels of anxiety at ICU admission are associated with death or the occurrence of one or more organ failure during the first 7 days, and that anxiety should be considered as a "warning sign" in critically ill patients.

Multicentre prospective study. Demographic data, diagnosis, severity scores and vital signs will be assessed at admission.

Within the first 12 hours, and at day 4 and day 7, anxiety will be assessed by the STAI scale, Geriatric Anxiety Inventory (GAI) scale, and a specific questionnaire assessing anxiety in the intensive care (REAX). Visual Analogue Scale (VAS) scoring for anxiety, starvation, thirst, pain and dyspnoea will be collected daily. Anxiety and depression will be evaluated by the Hospital Anxiety and Depression Scale (HADS) on day 7. Patients will be assessed by psychologists or trained doctors, the team in charge of the patient will not be aware of the result of this evaluation.

Clinical and biological data regarding the hemodynamic, neurological, respiratory and renal status will be collected from admission to day 7, as well as parameters required for computing the Sequential Organ Failure Assessment (SOFA) score.

Inclusion Criteria:

• All adult patients admitted to the ICU
• Patient without delirium at admission (assessed using the CAM-ICU score)

Exclusion Criteria:

• Severe hearing impairment or poor understanding of French language
• History of serious psychiatric illness
• Comatose or sedated patient at admission
• Diagnosis of voluntary self-poisoning, or polytrauma at admission

ICU patients, not comatose and not confused