Use of a ventilatory support system (BiPAP) for acute respiratory failure in the emergency department

Study objectives: Bilevel pressure ventilation has bad proven success in th e treatment of acute respiratory failure (ARF). The purpose of this study w as to identify patient characteristics early in the course of acute illness that can predict the successful use of bilevel pressure ventilation. Methods: Ventilatory assistance using a ventilatory support system (BiPAP m odel ST-D; Respironics; Murrysville, PA) was considered a treatment option for stable patients with ABF. The system was titrated to patient comfort. O nce stable settings had been achieved for 30 min, a posttrial arterial bloo d gas (ABG) measurement was obtained. Patient charts were reviewed for pret rial :Ind posttrial ABG levels, along with demographics, APACHE (acute phys iology and chronic health evaluation) II score, Glasgow Coma Scale (GCS), a nd length of stay (LOS) data. Results: Bilevel pressure ventilation trials were performed on 58 patients. In 43 patients (74.1%), the trials were successful. Of the 15 patients (25 .9%) in whom the trials were not successful, 13 patients required intubatio n. The pretrial ABG levels did not predict success, as there were no signif icant differences between the success and failure groups for pH and Pa-CO2, respectively: 7.26 vs 7.26 mm Hg and 75.3 vs 72.8 mm Hg. After 30 min, pos ttrial ABG levels for pH and Pa-CO2 predicted successful avoidance of intub ation: 7.34 vs 7.27 mm Hg (p < 0.002) and 61.9 vs 73.0 mm Hg (p < 0.04), re spectively. There were no significant differences between the success and f ailure groups in age, gender, GCS, or APACHE II. There were differences bet ween the success and failure groups for LOS data (ventilator days, ICU days , and hospital days): 1.8 vs 10.4 days (p < 0.01), 4.2 vs 12.3 days (p < 0. 02), and 7.5 vs 15.6 days (p < 0.02), respectively, Conclusion: Successful treatment with bilevel pressure ventilation could no t be predicted by pretrial data (including pH and Pa-CO2) obtained in the e mergency department; however, a successful outcome could be determined quic kly with a 30-min trial. Successful treatment with bilevel pressure ventila tion significantly reduced LOS data. Clinical implications: Our inability to predict success based on initial da ta supports the use of bilevel pressure ventilation trials for all stable p atients with ARF. If the patient's condition fails to improve within 30 min , intubation amd mechanical ventilation is indicated.