Test-Retest Reliability of Serial Pure-Tone Audiograms in Children at a School for the Deaf

Abstract

Successive air-conduction audiograms obtained between ages 1¼ and 12 years and recorded in the charts of 75 children at a school for the deaf were analyzed statistically. Approximately half of 414 audiogram pairs differed by at least 20 dB for 2 or more frequencies, and 34% of 3203 paired frequency scores differed by at least 20 dB. Conductive losses, moderate hearing losses, and perhaps mild mental retardation were associated with unreliability. Audiograms performed at the school by a nonprofessional audiologist were especially unreliable. The desirability of retesting a child until a stable threshold is obtained and of indicating the zone of uncertainty around each threshold point was stressed. When the zone of uncertainty is wide, a verbal label describing the estimated severity of the loss might be preferable to possibly invalid numerical information.