A comparison of three variable performance devices for postoperative oxygen therapy.

Eleven patients in the intensive care unit following major abdominal surgery with a nasogastric tube in situ and receiving oxygen via facemask were allocated to receive in a random sequence oxygen at 4 l.min-1 via a Hudson mask, nasal cannulae or a nasal catheter with foam collar at the distal end. A significantly greater PaO2 was achieved using both the nasal catheter with foam collar (p < 0.01) and Hudson mask (p < 0.05) compared to the nasal cannulae. There were no significant differences in the PaCO2 or SpO2 values obtained between the devices. In the majority of postoperative patients in whom a variable performance device is indicated, nasal devices appear preferable in terms of patient comfort and compliance. The nasal catheter with foam collar produced a significantly greater PaO2 than nasal cannulae in patients with a nasogastric tube in situ.