Non-invasive pressure support ventilation in acute hypoxemic (non hypercapnic) respiratory failure. Observations in Respiratory Intermediate Intensive Care Unit.

BACKGROUND: Non-invasive positive pressure support ventilation (NIPSV). METHODS: In patients with acute hypoxaemic (PaO2/FiO2 M100) non hypercapnic respiratory failure (ARF) admitted to a Respiratory Inter-mediate Intensive Care Unit of a general Hospital, between January 1993 and December 1997. RESULTS: In 21 selected patients (PaO2/ FiO2T0=82+/-9) NIPSV improved PaO2 in 13/21 patients (Group A) and did not improve in 8/21 patients (Group B) (PaO2/FiO2T1=154+/-25 in Group A vs PaO2/FiO2T1=106+/-7.5 in Group B, p=0.00001). Upon admission the two groups did neither significantly differ for blood gas values (PaO2/FiO2T0=84+/-9.6 in Group A vs 79.8+/-8.7 in Group B), nor for clinical status (APACHE II=19.8+/-5 in Group A vs 24.6+/-7 in Group B). Shorter duration of NIPSV in Group B patients (11.2+/-19.7 hrs vs 35.3+/-32.3 hrs in Group A, p=0.047), in spite of a rise in PEEP (9.3+/-2.3 in Group B vs 5.5+/-2.4 in Group A, p=0.003) and Pressure Support (18.7+/-1.8 in Group B vs 15+/-3.2 in Group A, p=0.004) was due to onset of conditions which required shifting from NIPSV to endotracheal intubation (ETI). OUTCOME: 8/21 patients were successfully treated by only NIPSV. 8/21 patients were intubated. 5/21 patients dead in RIICU; 1 month survival: 9/21 patients. Side effects: mask intolerance (3/21); skin necrosis (1/21); pneumothorax (1/21). CONCLUSIONS: NIPSV may be tried in ARF patients to improve PaO2 and avoid ETI.