Factors associated with noninvasive ventilation failure in postoperative acute respiratory insufficiency: an observational study

Abstract

Few data are available on the efficacy of noninvasive ventilation (NIV) in postoperative patients with acute respiratory failure (ARF). Seventy-two patients coming from the surgical wards with postoperative ARF were retrospectively evaluated. The major characteristics of patients who were intubated were compared with the characteristics of those who were not after a trial of NIV. Predictive factors for failure of NIV were analysed. Out of 72 patients with ARF after surgery who were treated with NIV, 42 avoided intubation (58%). On a univariate analysis, a decrease in the paO2/FiO2 ratio after 1 h of NIV (223 +/- 84 to 160 +/- 68 mmHg, P < 0.05) was associated with NIV failure and need for tracheal intubation because of nosocomial pneumonia and an increased simplified acute physiology score (SAPS) 2. In a multivariate analysis, nosocomial pneumonia [odds ratio (OR) 4.189; 95% confidence interval (CI) 1.383-12.687] and SAPS 2 higher than 35 (OR 4.969; 95% CI 1.627-15.172) were independent predictive factors of NIV failure. NIV success was associated with a reduced ICU stay (16.8 vs. 26.1 days, P < 0.001). NIV could be considered in postoperative patients who presented with ARF. Nosocomial pneumonia is predictive of NIV failure.