Noninvasive ventilation in weaning patients with hypoxemic acute respiratory failure (h-ARF): a multicentre randomized controlled study.

INTERVENTION: Non invasive ventilation (NIV) will be applied through a helmet, which is a transparent latex‐free polyvinylchloride hood, or by mask. The ventilator will be set with the same positive end expiratory pressure (PEEP) and inspiratory support applied to the patient before extubation. During the first 24 hours, NIV will be maintained continuously. Thereafter brief discontinuation (less than 5min) will be tolerated for nursing the patients (face, oral hygiene) or to change the non invasive interface. To enhance patient tolerance to NIV, a mild sedation with low doses of remifentanyl (0.025‐0.15 microg/kg/min) and/or clonidine (bolus 1 microg/kg/min, than 0.003‐0.01 microg/kg/min, max 750‐900 microg/die) and/or propofol (10‐50 mg/hr) will be allowed. Alternation of interfaces (helmet and mask) will also be accepted during the study period to increase patient tolerance. PEEP will be reduced by 2 cmH2O per hour whenever the PaO2/FiO2 ratio will exceed 225, to a minimum level of 8 cmH2O. No decrease in PEEP will be undertaken if PaO2/FiO2 ratio is below 225. In this case, the attendant physician will re‐evaluate the patient after 6 hours. Inspiratory support will be reduced by 2 cmH2O per hour to maintain PaCO2 <= 50 mmHg and pH >= 7.35, to a minimum level of 10 cmH2O. When NIV is set at the minimum levels of PEEP and inspiratory support (8 cmH2O and 10 cmH2O, respectively) and the PaO2/FiO2 is higher than 250, a 30‐minute spontaneous breathing trial (SBT) with a Venturi mask 35% will be attempted. The SBT will be interrupted and NIV resumed in case of severe dyspnea, activation of accessory muscles, and paradoxical abdominal motion, respiratory rate (RR) > 30 breaths/min, SpO2 < 90%. At the end of the trial an arterial blood gas analysis will be performed: the patients will be considered successfully weaned off NIV if pH >= 7.35, PaCO2 <= 50 mmHg and PaO2 >= 70 mmHg. The total duration of NIV treatment will depend on how quick, according to the established algorithm, will be the weaning process and it can vary on a case by case basis. CONDITION: Weaning in Hypoxemic Acute Respiratory Failure PRIMARY OUTCOME: Aim of the study is to investigate whether in patients with hypoxemic ARF early extubation followed by immediate application of NIV (interventional group) reduces the overall number of days spent in ICU. SECONDARY OUTCOME: Duration (days) of continuous intravenous sedation (monitored by clinicians). Hospital mortality, monitored by clinicians. ICU mortality, monitored by clinicians. Number of tracheotomy, monitored by clinicians. Side effects / complications of invasive mechanical ventilation: infection, sepsis and ventilator associated pneumonia (VAP), monitored by clinicians. INCLUSION CRITERIA: Invasive mechanical ventilation duration > 48 hours; PaO2/FiO2 ranging between 200 and 300 with a positive end‐expiratory pressure PEEP > 8 and < = 12 cmH2O and a FiO2 <= 0.6; Pressure Support Ventilation (PSV) with a total applied pressure (i.e. PEEP + inspiratory support) <= 25 cmH2O; PaCO2 <= 50 mmHg; pH >= 7.35; Respiratory rate (RR) <= 30/min; Core body temperature <= 38.5 (degrees celsius); Glasgow Coma Scale (GCS) >= 10; Presence of clearly audible cough during suctioning; Tracheal suctioning<= 2/hr