Little is known about the procedure of extubation of patients admitted in Intensive Care Units (ICU). In particular, effects of tracheal suction during extubation have never been evaluated. Tracheal suction induces alveolar derecruitment in sedated patients under mechanical ventilation and is a major source of pain.

The aim of this study was to evaluate the impact of tracheal suction during the extubation procedure of critically ill patients on the end-expiratory lung volume.

This is a prospective, monocentric study, conducted in the surgical ICU of the university hospital of Rouen, France.

Sixty patients were expected to be randomized before extubation into two groups (ratio of 1:1) with different extubation protocols depending on whether tracheal suction was performed or not.

After oral information and collection of the non opposition of the patient to participate in the study, eligible patients were randomized (raio 1:1) in two groups: "tracheal suction" group or "no tracheal suction" group.

The allocation concealment was assured by enclosing assignments in sequentially numbered, opaque, sealed envelopes. Envelopes were opened after enrolment of each patient by the medical doctor in charged. Each envelope contained a number by a random allocation process using a computer-generated random block design (the randomization list was established by the local biostatistics unit before the beginning of the study).

Juste after inclusion, the 30 minutes standardized extubation protocol started and consisted of:

• arterial blood gas analysis before the extubation (if there wasn't one dating less than 6 hours),
• adjustment of the backrest of the bed in tilt to + 45 °,
• tracheal suction 30 minutes before extubation (using a 14 french catheter, a vacuum of -200 mmHg systematically measured by a manometer XX),
• the ventilator was then set on pressure support ventilation with pressure support level of 8 cmH2O and positive end-expiratory pressure (PEEP) of 5 cmH2O (FiO2 was adjusted for oxygen saturation by pulse oximetry between 95 and 98%) for 30 minutes,
• installation of electrode belt for electrical impedance tomography (EIT) monitoring (Pulmovista 500, Dräger®) and calibration of the system,
• aspiration of oropharyngeal secretions immediately before extubation with an oral cannula.
• for "tracheal suction" group, extubation occured 30 minutes after inclusion. A tracheal suction (using a 14 french catheter, a vacuum of -200 mmHg) was performed at the same time as removal of the tracheal tube, after disconnection of the ventilator and after deflating the balloon of the tracheal tube.
• for "no tracheal suction" group, extubation occured 30 minutes after inclusion and was performed after deflation of the balloon (and without further maneuver).
• all patients underwent chest physical therapy between the 15th and 60th minutes following extubation.

No calculation of the number of subjects needed was possible (no data available concerning ΔEELI at extubation).

Data were described in the whole population and for each group ("tracheal suction" and "no tracheal suction") using the usual descriptive parameters: frequency for qualitative variables, median and interquartile range (IQR) for quantitative variables. Statistical analysis consisted of a nonparametric Mann and Whitney test for the quantitative variables and an exact Fisher test for the qualitative variables (using Statistical Analysis System software, version 9.4, Statistical Analysis System Institute; Cary, NC). The significance of the tests was retained for an α risk of 5%.

Inclusion Criteria:

• age of 18 years or more
• hospitalization in the surgical ICU (whatever the cause of hospitalization)
• under mechanical ventilation via a tracheal tube (oro or nasotracheal) for at least 24 hours
• satisfying general criteria for mechanical ventilation weaning (described by the French Language Resuscitation Society)
• having successfully completed a spontaneous breathing trial (among those described by the SRLF)
• physiotherapist available during the first hour after extubation

Exclusion Criteria:

• the presence of an electrical implantable medical device (pacemaker, automatic defibrillator, deep brain stimulation box)
• body mass index (BMI) > 50
• pregnancy
• tracheal tube with subglottic suction channel
• technical impossibility of monitoring by electrical impedance tomography (chest plaster, undrained pneumothorax, ...).