Non-invasive ventilation as a first-line treatment for acute respiratory failure: "real life" experience in the emergency department.

OBJECTIVE To describe our experience with non-invasive ventilation (NIV) for patients with acute respiratory failure (ARF) in the emergency department (ED). METHODS A prospective/retrospective, observational study on 190 patients with ARF (mean +/-SD age 72.2+/-12.9 years, mean APACHE II score 18.9+/-5.9), who received 200 NIV trials in an ED. We analysed the NIV register data (prospectively collected) and medical records (retrospective data abstraction) and evaluated clinical indications for NIV, patient outcomes, and predictive factors for success and death. NIV success was defined as tolerance of the procedure and no need for endotracheal intubation (ETI). RESULTS Main indications to NIV were cardiogenic pulmonary oedema (CPE) (70 trials), acute exacerbation of COPD (39), both CPE and acute exacerbation of COPD (11), pneumonia (48), decompensation of obesity/hypoventilation (6), other conditions (26). The procedure was successful in 60.5% of trials. Global mortality was 34.5%, similar to the APACHE II predicted mortality of 32%. ETI rates were 6.5% and tracheostomy rates 1%. The improvement of pH within six hours after NIV initiation was predictive of survival in the hypercapnic group. CONCLUSIONS Our results confirm the global efficacy of NIV in an ED setting, and show that, in spite of lower success rate in "real practice" in comparison with RCTs, an intermediate care unit can represent an appropriate and less expensive setting to perform this technique. The low rate of ETI seems to be because of the high number of patients for whom NIV was used as "ceiling" treatment.