RESUMO
OBJECTIVE: To assess the effectiveness and identify predictors of failure of noninvasive ventilation. DESIGN: A retrospective, longitudinal descriptive study was made. SETTING: Adult patients with acute respiratory failure. PATIENTS: A total of 410 consecutive patients with noninvasive ventilation treated in an Intensive Care Unit of a tertiary university hospital from 2006 to 2011. PROCEDURES: Noninvasive ventilation. MAIN VARIABLES OF INTEREST: Demographic variables and clinical and laboratory test parameters at the start and two hours after the start of noninvasive ventilation. Evolution during admission to the Unit and until hospital discharge. RESULTS: The failure rate was 50%, with an overall mortality rate of 33%. A total of 156 patients had hypoxemic respiratory failure, 87 postextubation respiratory failure, 78 exacerbation of chronic obstructive pulmonary disease, 61 hypercapnic respiratory failure without chronic obstructive pulmonary disease, and 28 had acute pulmonary edema. The failure rates were 74%, 54%, 27%, 31% and 21%, respectively. The etiology of respiratory failure, serum bilirubin at the start, APACHEII score, radiological findings, the need for sedation to tolerate noninvasive ventilation, changes in level of consciousness, PaO2/FIO2 ratio, respiratory rate and heart rate from the start and two hours after the start of noninvasive ventilation were independently associated to failure. CONCLUSIONS: The effectiveness of noninvasive ventilation varies according to the etiology of respiratory failure. Its use in hypoxemic respiratory failure and postextubation respiratory failure should be assessed individually. Predictors of failure could be useful to prevent delayed intubation.
