The purpose of this study is to determine whether drainage of subglottic secretions is effective to reduce the incidence of pneumonia acquired under mechanical ventilation in ICU's patients.

Nosocomial infection represent a major problem in hospitals. In intensive care units, ventilator-associated pneumonia (VAP) is associated with an increase in morbidity and mortality. Incidence of VAP is high, 20 to 30% of patients under mechanical ventilation (MV) for more than 48 hours. In the pathogenesis of VAP, two processes are considered essential for its development: bacterial colonization of the oropharynx and tracheobronchial tract, followed by aspiration of contaminated secretions into the lower airways. It has been shown that pooled secretions above inflated endotracheal tube cuffs may be source of aspiration and can be a cause of VAP. Several studies have suggested that recurrent aspiration of subglottic secretions can be prevented by intermittent drainage of subglottic secretions. Nevertheless, recent CDC guideline reported that there was not enough evidence to conclude on the efficacy of such intervention.

During the study, patients, in both groups, will be intubated with the same device permitting subglottic drainage (HiLo Evac endotracheal tube). The daily screening of VAP will be performed until the 28th days of mechanical ventilation without occurrence of VAP. The follow-up of each patient will be realized until the ICU's discharge.

Comparison: The incidence of VAP will be compared between two groups: one group with intermittent drainage of subglottic secretions and the other one without this intervention.

Inclusion Criteria:

• ICU patients expected to require MV more than 48 hours

Exclusion Criteria:

• patients already intubated for more than 12 hours
• patients intubated with another tube than the HiLo Evac tube
• patients admitted for cardiac arrest
• patients admitted for self poisoning
• patients admitted with tracheotomy
• patients already included in an another clinical trial