Severe diseases in intensive care unit (ICU) patients are associated with a high mortality rate which nevertheless remains difficult to predict. Recently, the abolition of some brainstem reflexes at clinical examination of ICU patients within the first 24 hours after has been shown of prognostic value in ICU patients requiring sedation. Early abolition of the cough reflex was associated with an increase in mortality and that of the oculocephalic reflex was predictive of coma or delirium after sedation has been stopped. A dysfunction of the brainstem may account for these results and be present in other patient subpopulations, particularly those who do not receive iv sedation or the brain injured patients, who were eluded from the previous study. This dysfunction could take place in the muti-organ failure characteristic of the severe ICU patient. On the other hand, a preliminary study performed on somatosensory evoked potentials has shown that a latency of the P14 wave greater than 16 ms between day 1 and day 3 after admission was associated with death at 28 days. The primary goal of this project is to confirm this hypothesis. By studying the clinical and electrophysiological responses of the brainstem in ICU patients, with or without brain injury, with alteration of consciousness in relation or not with sedation. The other objectives are to determine the correlations between neurophysiological clinical neurological, or neuroradiological data with delirium occurrence.

The main objective of this study is to determine, in ICU patients with or without brain injury and with alteration of consciousness in relation or not with sedation, if abolition of the cough reflex at Day 1 after admission is predictive of mortality at 28 days, independent from cause for admission and severity.

Adult Medical or surgical, brain-injured or not, ICU patients requiring invasive mechanical ventilation for at least 48 hours and with alteration of consciousness induced or not by sedatives. Inclusion will take place at the 24th hour (± 12) after admission into the ICU will be enrolled. Those patients with either pregnancy, post anoxic coma, brain death, pre-existing neurologic disease disturbing the interpretation of the brainstem reflexes (Guillain-Barre, myasthenia, gravis, brain tumor, inflammatory or degenerating disease of the posterior fossa, acute peripheral neurologic disease), or declined participation will be excluded.

The measurements will be based on:

D1 : Demographic data, cause for admission, brainstem reflexes, Glasgow Coma Scale score, Confusion Assessment Method for the ICU (CAM-ICU) , Richmond Agitation Sedation Scale (RASS), Behavioral Pain Scale (BPS), simplified index of gravity (IGS 2) score, Sequential Organ Failure Assessment (SOFA) scores, cumulated sedative doses and Secondary Systemic Cerebral Aggressions (ACSOS) parameters D3 : parameters of D1, electrophysiologic analyses: Electroencephalogram (EEG), auditory evoked potentials (EAEP), Somatosensory Evoked Potential (SEP), recording of clinico-biological data, Cognitive evoked potentials (CEP) in patients with consciousness alteration still on mechanical ventilation at day 3.

D1-D28 : date of death, date of extubation, SOFA, CAM-ICU, cumulated doses of sedatives and analgesics, duration of sedation, occurrence of nosocomial pneumonia after Day 2, brain imaging data if available.

D28 : Glasgow Outcome Coma Scale and mini mental state

Inclusion Criteria:

• Adult Medical or surgical, brain-injured or not, patients
• Admitted to the intensive care Unit (ICU)
• Requiring invasive mechanical ventilation for at least 48 hours
• With alteration of consciousness induced by sedatives or not
• Within the 24 hours (± 12) after admission into the ICU

Exclusion Criteria:

• Pregnancy
• Post anoxic coma
• Brain death
• Pre-existing neurologic disease disturbing the interpretation of the brainstem reflexes (Guillain-Barre, myasthenia, gravis, brain tumor, inflammatory or degenerating disease of the posterior fossa, acute peripheral neurologic disease)
• Declined participation