OBJECTIVES: To determine the factors associated with failure of noninvasive positive pressure ventilation (NPPV) in patients presenting with acute respiratory failure to the emergency department (ED). METHODS: The authors retrospectively analyzed patients admitted to the ED for acute respiratory failure (defined as a PaCO2 level >45 mm Hg, and pH < or = 7.35 or a PaO2/FiO2 ratio < 250 mm Hg) and who were treated with NPPV. NPPV was delivered routinely according to an institutional protocol. Failure of NPPV was defined as the requirement of endotracheal intubation at any time. RESULTS: A total of 104 patients were included. NPPV failed in 31% (32/104), and the mortality was significantly higher in this group (12/32 [44%]) compared with patients who were not intubated (2/72 [3%]) (p < 0.0001). Factors associated with failure of NPPV were Glasgow Coma Scale score < 13 at ED admission (odds ratio [OR], 3.67; 95% confidence interval [CI] = 1.33 to 10.07), pH < or = 7.35 (OR, 3.23; 95% CI = 1.25 to 8.31), and respiratory rate (RR) > or =20 min(-1) (OR, 3.86; 95% CI = 1.33 to 11.17) after one hour of NPPV. The negative predictive value for NPPV failure was 86% (95% CI = 70% to 95%) for RR > or =20 min(-1). In the multivariate analysis, pH < or = 7.35 and RR > or =20 min(-1) after one hour of NPPV were independently associated with NPPV failure (adjusted ORs, 3.51; 95% CI = 1.29 to 9.62 and 3.55; 95% CI = 1.13 to 11.20, respectively). CONCLUSIONS: Patients with pH < or = 7.35 and an RR > or =20 min(-1) after one hour of NPPV had an increased risk of subsequent endotracheal intubation.
OBJECTIVES: To determine the factors associated with failure of noninvasive positive pressure ventilation (NPPV) in patients presenting with acute respiratory failure to the emergency department (ED). METHODS: The authors retrospectively analyzed patients admitted to the ED for acute respiratory failure (defined as a PaCO2 level >45 mm Hg, and pH < or = 7.35 or a PaO2/FiO2 ratio < 250 mm Hg) and who were treated with NPPV. NPPV was delivered routinely according to an institutional protocol. Failure of NPPV was defined as the requirement of endotracheal intubation at any time. RESULTS: A total of 104 patients were included. NPPV failed in 31% (32/104), and the mortality was significantly higher in this group (12/32 [44%]) compared with patients who were not intubated (2/72 [3%]) (p < 0.0001). Factors associated with failure of NPPV were Glasgow Coma Scale score < 13 at ED admission (odds ratio [OR], 3.67; 95% confidence interval [CI] = 1.33 to 10.07), pH < or = 7.35 (OR, 3.23; 95% CI = 1.25 to 8.31), and respiratory rate (RR) > or =20 min(-1) (OR, 3.86; 95% CI = 1.33 to 11.17) after one hour of NPPV. The negative predictive value for NPPV failure was 86% (95% CI = 70% to 95%) for RR > or =20 min(-1). In the multivariate analysis, pH < or = 7.35 and RR > or =20 min(-1) after one hour of NPPV were independently associated with NPPV failure (adjusted ORs, 3.51; 95% CI = 1.29 to 9.62 and 3.55; 95% CI = 1.13 to 11.20, respectively). CONCLUSIONS:Patients with pH < or = 7.35 and an RR > or =20 min(-1) after one hour of NPPV had an increased risk of subsequent endotracheal intubation.