Background: Prolonged mechanical ventilation (MV) exposes patients to increased risk of infection and mortality. A weaning protocol (WP) reduces the duration of MV. However, these protocols remain complicated to implement in clinical practice. These considerations led to the development of automated weaning systems (AWS) which are expensive and limited to expert centers.

Scope of the problem: AWS are not yet the standard care because evidence of AWS benefit is weak and several unanswered questions remain regarding their roles in weaning. There is a need for building new ways to improve WP. Indeed, medical staff's compliance with a daily screening of patients to wean is low. Moreover, modifications in ventilator settings require the availability of physicians at bedside and are rarely modified on a daily basis. This leads to impaired weaning process and prolonged MV for patients.

Acknowledge gap: Capnography (CAP) has been validated for intubation procedures and integrated into AWS. Some studies encouraging use of CAP have shown an optimization of MV, a decrease of unnecessary arterial blood gas without impacting patients' prognosis. However, the role of CAP in WP in intensive care units (ICU) has not been assessed so far.

Hypothesis: A nurse driven WP based on CAP may improve accuracy of weaning procedures at bedside. Such protocol might be easy to apply in daily practice and could reduce both the duration of MV and its related complications.

Methods: This is a prospective, bi-center, pilot study to assess the feasibility of a new nurses managed WP based on capnography among patients requiring more of 48 hours of MV. Enrollment of 100 consecutive patients in two ICU of the universitary hospital in Reims. The primary endpoint is the protocol compliance. Secondary endpoints are: capacity to performed a weaning (time to first extubation, total duration of MV...) caregivers' acceptance and feelings regarding the protocol, a safety evaluation with report of hospital mortality, ventilator associated-pneumonia, autoextubation, length of stay and the number of ABG/patient/day during the study period. Inclusion critera: all patients on MV for more than 48 hours and stabilization of the ICU entry pathology. Exclusion critera: Patients with severe neuropathy, cardiopathy or pulmonary disease. Weaning protocol begins when the following criteria are present: PEEP ≤ 8 cmH2O, PaO2/FiO2 ratio ≥150 (mmHg) or SaO2 ≥ 90% with FiO2 ≤ 0.5. The final criterion for study entry is successful completion of a 30-min spontaneous breathing test (SBT) using pressure support (PS) (max 30 cmH2O). After a successful test, nurse reduces PS by 2 mmHg every 3 hours and the FIO2 as long as the patient remains in his comfort zone defined as: FR 12-30/min, VT ≥ 6ml/kg and ETCO2 ≤ 55 mmHg (≤65 mmHg for chronic obstructive pulmonary disease (COPD) patients). A scale allows FIO2 and PEEP to be titrated with an oxygen saturation between 90% and 98% (92% for COPD). Nurses recorded every 3 hours ventilatory, hemodynamic, sedation parameters and modifications. At any time physicians can return to the previous settings if the patient is leaving the comfort zone. SBT are conducted when a minimum level of PS and PEEP is reached, 7cmH20 and 5cmH20 respectively. SBT performed using a T piece with or no oxygen for 30-120 min. If a patient passes a successful SBT and meets extubation criteria, he is extubed. Extubation and SBT protocols are performed according to French guidelines SFAR/SRLF. Nevertheless, extubation always remains the decision of the attending physician. Protocols continue until successful extubation (at least 48 hours of unassisted spontaneous breathing), ICU death, ICU discharge or until day 90 after randomization.

Interests and Scientific impacts: For patients, a potential reduction of VM duration and its associated-complications. For caregivers, a more efficient WP and a better optimization of workload. This pilot study will be a step before to investigate larger studies to compare our WP to standard care.

inclusion criteria :

• patients aged over 18 years and admitted in intensive care
• invasive ventilation with assisted mode (assist-control or spontaneous with pressure support) for more than 48 hours,
• stabilization of the cause of admission defined as: positive end-expiratory pressure level lower than or equal to 8 cm of water, a arterial oxygen saturation higher than 90 percent with a fraction of inspired oxygen lower than or equal to 50 percent for acute respiratory failure; infusions of norepinephrine less than or equal to 0,2 gamma/kg/min, no epinephrine or dobutamine for cardiogenic or septic shock; body temperature 36-38,5°C.

exclusion criteria :

• schedulded extubation within 24 hours after the inclusion,
• schedulded surgery within week after the inclusion,
• no available capnometer,
• end stage chronic disease (cardiac, pulmonary, or renal),
• "do not resuscitate order",
• prior ventilation greater than 24 hours during the same hospitalization,
• tracheostomy,
• known or suspected severe myopathy or neuropathy,
• severe chronic obstructive pulmonary disease,
• pregnancy,
• patients without medial care assistance.
• conditions and/or circumstances that might result in difficulty to understand the informed consent