Noninvasive pressure support ventilation (NIPSV) with face mask in patients with acute cardiogenic pulmonary edema (ACPE)

Objectives: To assess (1) the short-term hemodynamic, respiratory and arter ial blood gas ef fects of NIPSV in patients with ACPE who were likely to re quire endotracheal intubation, (2) the initial causes of failure and (3) th e side effects and the difficulties of this technique. Design: Uncontrolled, prospective clinical study. Setting: Teaching hospital intensive care unit. Patients: 26 consecutive patients with severe ACPE. Interventions: Noninvasive ventilation via a face mask, using a pressure su pport mode (20.5 +/- 4.7 cmH(2)O), with an initial fractional inspired oxyg en of 93.0 +/- 16 % and a positive end-expiratory pressure of 3.5 +/- 2.3 c mH(2)O. The need to intubate the patients within 48 h was considered as a c riterion of failure of the procedure. Measurements and results: Clinical and biological parameters were measured at 15 and 30 minutes, 1 h and 2 h and at 1 h and 2 hi respectively. There w ere 5 (21%) failures and 21 (79 %) successes. In both the success and the f ailure groups, clinical and blood gas parameters improved at the first meas ure. In the success group, within 15 min of the start of NIPSV, pulse oxime try saturation (SpO(2)) had increased from 84 +/- 12 to 96 +/- 4% (p < 0.00 1), the respiratory rate (RR) had decreased from 36 +/- 5.3 to 22.4 +/- 4.9 breaths/min (p < 0.0001) and within 1 h the arterial oxygen tension and pH , respectively, had increased from 61 +/- 14 to 270 +/- 126 mmHg (p < 0.000 1) and from 7.25 +/- 0.11 to 7.34 +/- 0.07 (p < 0.01) and the arterial carb on dioxide tension (PaCO2) had decreased from 54.2 +/- 15 to 43.3 +/- 6.4 m mHg (p < 0.01). There were no statistical differences between the success a nd failure groups for the initial clinical parameters: SpO(2), RR, heart ra te, mean arterial pressure. The only differences between the success and fa ilure groups were in the PaCO2 (54.2 +/- 15 vs 32 +/- 2.1 mmHg, p < 0.001) and the creatine kinase (CPK) (176 +/- 149 vs 1282 +/- 2080 IU/l, p < 0.05) ; this difference in CPK activity was related to the number of patients who had an acute myocardial infarction (AMI) (415 in the failure group vs 21 2 1 in the success group, p < 0.05). All patients with AMI in the failure gro up died. Conclusion: Among patients in acute respiratory failure, those with severe ACPE could benefit from NIPSV if they are hypercapnic, but NIPSV should be avoided in those with AMI.