Effect of spontaneous sighs on arterial oxygenation during isoflurane anesthesia in humans.

The presence, frequency, and volume of spontaneous sighs was evaluated in 21 (ASA 1-2) supine patients aged 44 +/- 15.2 (SD) yr, during isoflurane-nitrous oxide anesthesia. Before induction the inspiratory capacity of each patient was determined. After induction of anesthesia and tracheal intubation patients breathed spontaneously except for three manual inflations to each patient's predetermined inspiratory capacity at the beginning and end of surgery. Arterial blood gas tensions were measured before and 5 min after each set of mechanical deep breaths and each hour during surgery, the mean duration of which was 2 +/- 0.09 hr. Spontaneous sighs occurred in 13 of 21 patients. The average frequency was 6 +/- 4 sighs/hr. At FIO2 = 0.5, nonsighing patients had an initial PaO2 of 229 +/- 59 mm Hg and sighers had an initial PaO2 of 162 +/- 57 mm Hg (P less than 0.05). Arterial oxygen did not change in sighing patients during the course of surgery, while in nonsighing patients the PaO2 decreased from the initial value of 229 +/- 60 mm Hg to 170 +/- 63 mm Hg (P less than 0.05). Mechanical deep breaths administered at the end of surgery produced no improvement in oxygenation in either sighers or nonsighers. The presence or absence of sighs did not correlate with PaO2 or PACO2. Though the results suggest that spontaneous sighs in some patients may function to help maintain arterial oxygenation, all patients maintained their PaO2 while breathing spontaneously under general anesthesia in the supine position.