The use of sound recording and oxygen saturation in screening snorers for obstructive sleep apnoea.

It is desirable to screen snoring patients for obstructive sleep apnoea (OSA) prior to surgical treatment. We postulated that the addition of a sound profile would increase the value of overnight oxygen saturation (SaO2) as a screening method. Thirty-nine polysomnographic studies including sound level measured by calibrated meter were performed on snorers being considered for uvulopalato-pharyngoplasty (UPPP). Polysomnography showed an apnoea/hypopnoea index (AHI) > or = 15 per hour of sleep in seven subjects. Two experienced observers independently, without knowledge of other data, classified paper records of SaO2 alone and SaO2 plus sound level obtained during polysomnography as OSA 'unlikely', 'equivocal' or 'definite'. The addition of sound to SaO2 reduced the number of equivocal results from 14 to six and increased the number classified as 'definite' or 'unlikely'. The sensitivity of oximetry +/- sound increased as the threshold AHI used in the definition of OSA increased; addition of sound improved recognition of mild OSA without impairing specificity.