The clinical efficacy of noninvasive ventilation (NIV) has now been demonstrated in the management of acute-on-chronic respiratory failure (ACRF) of various etiologies. Endotracheal mechanical ventilation (ETMV) can lead to numerous complications and weaning difficulties increasing the risk of prolonged ETMV, morbidity and mortality as well as excess cost of intensive care. Therefore, it could be interesting to use NIV for delivering effective ventilatory support and reduce the length of ETMV in ACRF patients still not capable to maintain spontaneous breathing. From the interesting but discordant results of two recent randomised controlled trial, a working group from the Société de Réanimation de Langue Française (SRLF) decided to perform a new prospective randomised controlled and multicenter trial. The aim of the study is to assess the usefulness of NIV as an extubation and weaning technique in ventilated ACRF patients. The methodology used compares three weaning strategies in parallel in ACRF patients considered difficult to wean : invasive conventional weaning (group A), extubation relayed by nasal oxygentherapy (group B), and extubation relayed by NIV (group C). Based on the main end-point defined as the weaning success/failure rate, 208 patients from17 investigator centers are planned to be included. Results of the study will also allow to assess the respective impact of the three weaning strategies on the length of ETMV and weaning, the mechanical ventilation-related morbidity, the patients lengths of stay and mortality.

Results of the VENISE trial should permit to improve the management of the difficult to wean ACRF patients and thus to contribute to more define the place of NIV among the weaning and prevention of re-intubation strategies in these patients.

Inclusion Criteria:

• endotracheal mechanical ventilation for 48 hours or more for acute respiratory failure in chronic respiratory failure patients (obstructive or restrictive disease),
• patient ready to be weaned (steady respiratory and hemodynamic conditions for 24 hours),
• spontaneous breathing trial failure occurring between 5 minutes and 2 hours, judged on clinical and/or arterial blood gas tolerance,
• written informed consent obtained (patient or family)

Exclusion Criteria:

• respiratory and hemodynamic instability
• initial intubation considered as difficult
• swallowing disorders suspected
• inefficient cough
• bronchial hypersecretion at the weaning time
• non cooperant patient
• contra-indications for nasal or facial mask (facial skin lesions,...)
• recent history of upper gastro-intestinal surgery
• recent history of myocardial infarction