• Primary outcome: The incidence of respiratory failure after extubation [ Time Frame: 3 days after extubation ] [ Designated as safety issue: No ]
  

• Survival [ Time Frame: 90 days ] [ Designated as safety issue: No ]
  

Background and aims: Patients with chronic respiratory disorders, intubated and mechanically-ventilated in intensive care units (ICU), often exhibit hypercapnia during the spontaneous breathing trial prior to extubation; this is associated to increased incidence of extubation failure and mortality. We will assess the efficacy of noninvasive ventilation (NIV) in the prevention of extubation failure and mortality in these patients. Design: Prospective, randomized, controlled clinical trial with two groups: support with noninvasive ventilation, and conventional clinical management (control group). Subjects: Patients with chronic respiratory disorders, intubated and ventilated ³48 hours, with hypercapnia (PaCO2 >45 mmHg) during the spontaneous breathing trial, ready to be extubated. Interventions: Random allocation into two groups: NIV immediately after extubation during 24 hours, on conventional clinical management. Clinical follow-up (incidence of extubation failure, reintubation, ICU and hospital mortality, 90-day survival, complications associated to mechanical ventilation, ICU and hospital length of stay). Analyses of data: Comparison between two groups (NIV and control) in non-paired data design: quantitative variables (t-test, Mann-Whitney test, ROC curves), qualitative or categorical variables (chi square, Fisher's exact test), and Kaplan-Meier estimate of survival curves).