PURPOSE: The purpose of this study was to compare the clinical results of extracorporeal membrane oxygenation (ECMO) using a centrifugal pump (CP group) and that using a roller pump (RP group) for the treatment of acute respiratory failure (ARF). METHODS: From November 1990 to July 2001, the ECMO system was introduced for the treatment of 15 patients with ARF; 10 cases of pneumonia or adult respiratory distress syndrome (ARDS), 4 cases of pulmonary hemorrhage, and 1 case of hypoxemia following cardiac surgery. Five patients were included in the RP group, and 10 were included in the CP group. RESULTS: The mean PaO2/FIO2 ratio, PaCO2 prior to ECMO induction, and the mean duration of ECMO support were 59.8, 38.8 mmHg, and 125 h, respectively, in the RP group, and 65.6, 82.0 mmHg, and 107 h, respectively, in the CP group. The mean PaCO2 value was significantly (P < 0.05) higher in the CP group than in the RP group due to the ventilation with "permissive hypercapnia." In the RP group, all patients died of either complications or recurrence due to ARF. In the CP group, 3 patients including 2 with pulmonary hemorrhaging were discharged. CONCLUSIONS: Although the centrifugal pump ECMO improved the clinical results of ARF, the overall prognosis nevertheless depended on the original disease.
PURPOSE: The purpose of this study was to compare the clinical results of extracorporeal membrane oxygenation (ECMO) using a centrifugal pump (CP group) and that using a roller pump (RP group) for the treatment of acute respiratory failure (ARF). METHODS: From November 1990 to July 2001, the ECMO system was introduced for the treatment of 15 patients with ARF; 10 cases of pneumonia or adult respiratory distress syndrome (ARDS), 4 cases of pulmonary hemorrhage, and 1 case of hypoxemia following cardiac surgery. Five patients were included in the RP group, and 10 were included in the CP group. RESULTS: The mean PaO2/FIO2 ratio, PaCO2 prior to ECMO induction, and the mean duration of ECMO support were 59.8, 38.8 mmHg, and 125 h, respectively, in the RP group, and 65.6, 82.0 mmHg, and 107 h, respectively, in the CP group. The mean PaCO2 value was significantly (P < 0.05) higher in the CP group than in the RP group due to the ventilation with "permissive hypercapnia." In the RP group, all patients died of either complications or recurrence due to ARF. In the CP group, 3 patients including 2 with pulmonary hemorrhaging were discharged. CONCLUSIONS: Although the centrifugal pump ECMO improved the clinical results of ARF, the overall prognosis nevertheless depended on the original disease.