A gradual increase in the volume of the induced gas space follows artificial pneumoperitoneum in dogs and cats with sulphur hexafluoride (SF,).' The volume change occurs because the rate of diffusion of nitrogen from the blood in to the pneumoperitoneum exceeds the rate of loss of SF, from the pneumoperitoneum to blood. The volume increases in artificial pneumothoraces during nitrous oxide anaesthesia are well-known.2 In these cases the rate of diffusion of nitrous oxide inwards exceeds the rate of diffusion of nitrogen outwards. The effect of similar anaesthesia on a non-compliant gas space causes rapid increases in pressure in the air-filled cisternal spaces of dogs following N 2 0 admini~tration.~ Similar effects occur in the middle ear during N 2 0 anaesthesia in patients4 Measurements in middle ear gas pressure and composition in cats and in man have confirmed these finding^.^." The effect is independent of airway pressure, and temporary or permanent hearing loss has occurred in several patients following N 2 0 anaesthesia.'. * Tympanic membrane graft displacement is commonly seen in middle ear surgery under N20 anaesthesia, yet the agent is used almost universally in this work. The possible effects of N 2 0 excretion into the middle ear cavity suggest that the gas should be withdrawn some minutes before the middle ear is ~l o s e d .~*~-" It has also been suggested that N 2 0 anaesthesia may be hazardous for the hearing of patients with polyethylene struts or surgical reconstructions for chronic ear disease.8The present study was undertaken because of the occurrence of a marked hearing loss following N 2 0 anaesthesia in one such patient. Its purpose was to confirm the effects of the gas on the normal middle ear, to compare the effect with that of non-N20 anaesthesia and to explore the time course of the events observed. Case historyThe patient, a 43-year-old woman, had a left stapedectomy for otosclerosis about 10 years ago and a standard 4.5-mm Schucknecht gelfoam and stainless steel prosthesis was inserted. Figure l(a) shows her audiogram at postoperative review; the upper trace is her (normal) sensorineural level and the lower is her hearing level to air conduction. The functional result of surgery was good with complete closure of the conductive gap.A year later she had a hysterectomy under general anaesthesia and immediately after operation complained of severe deterioration in her hearing. Her complaint was confirmed by audiometry which showed a 20-25 dB conductive loss- Figure I(b)-which has persisted to the present. The anaesthetic sequence was thiopentone, alcuronium, intubation and ventilation with N 2 0 and O2 to which 0.5% halothane was added.

No abstract