The purpose of this study is to evaluate the impact of an optimized suctioning procedure on the risk of endotracheal microaspiration, measured by tracheal to oropharyngeal ratio of amylase enzymatic activity.

• Study design : This is an open prospective, cross-over cohort study, including patients ventilated in one medical intensive care unit. Two procedures will be compared in two distinct, randomly decided cross-over consecutive periods

• Detailed description : Standard cares include tracheal suctions several times a day. They can induce a cough reflex in non paralysed patients leading to the mobilisation of the endotracheal tube and a consecutive raise in the risk of tracheal microaspiration. Thus drainage of subglottic secretion before tracheal suction is expected to reduce microaspiration. Subglottic secretion drainage could decrease the risk of Ventilator Associated Pneumonia even if it remains to be confirmed. However, no study assessed its efficiency in reducing tracheal microaspiration.

An optimization of the oropharyngeal suction procedure will include the use of a subglottic drainage in a specified order to realize a so called "optimized oral care". A comparison will be done with a routine oral care.

We will conduct a crossover study in which the patient's follow-up will last 2 days. Length of suction procedure and mouth care, volume of oral, subglottic and tracheal secretions and their qualitative appearance will be collected. The amylase enzymatic activity will be assessed in oropharyngeal, subglottic and tracheal samples.

• Registry procedure : All requested information will be recorded on case report forms and explanation will be reported for any missing data. False data will be clearly corrected, and signed by investigator or authorised person. For biochemical data (amylase enzymatic activity), results will be transmitted through excel software via a key and data will be integrated to the principal excel file on professional computer of the investigator.

Monitoring will be realised by the co-investigator in association with the clinical research associate.

• Quality factors. This clinical research associates will regularly check data to compare data entered into the registry against predefined rules for range or consistency with other data fields in the registry and a report will be report after each checking.
• Sample size assessment: Number of subjects : 24 patients. Six months should be necessary to include these patients.
• Statistical analysis : The sample size estimate assumed a median ratio of 8% and an estimated reduction from 8 to 2% with the optimized oral care. We determined that 12 patients per group would provide a power of 80%, with an α level of 5%.

Results will be expressed as means ± Standard Deviation. Continuous variables will be compared using the Student's t-test for matched paired series and categorical variables using the Chi 2 test. All p values and confidence intervals will be two-tailed. p<0.05 will be considered to be statistically significant.

Inclusion Criteria:

• Intubated with a subglottic secretion drainage device
• ventilated more than 48 hours
• patients yielding a significant tracheal suctioning volume (on a semi-qualitative evaluation: ≥ ++ on a scale consisting of 0,+,++,+++)

Exclusion Criteria:

• paralysed patients
• patients with Ramsay 1, 2
• patients breathing spontaneously
• patients less than 18 years old
• patients in a moribund state
• contra-indication for suctioning procedure (risk of haemorrhage, bronchospasm, Acute Respiratory Distress Syndrome with P/F < 100, malformation or tracheal fistula)
• bronchiectasis, cystic fibrosis, chronic obstructive pulmonary disease
• patients necessitating a closed-suctioning device (prevention of respiratory multiresistant bacteria transmission, Acute Respiratory Distress Syndrome with P/F ratio < 100)
• pregnant women