[Brain stem audiometry and unconventional audiometry in small children--a comparison with pure-tone audiometry performed at a later time].

Auditory brain-stem responses (ABR) was performed in the determination of hearing threshold in 71 infants and children born 1980-1985, all belonging to high risk groups or difficult-to-test children (median age 17 months, range 2-80). In a follow up study, 41 of the children were tested with pure tone audiometry approximately 2 1/2 years later. A significant correlation was revealed between the threshold obtained by ABR and the pure tone threshold at 2 kHz (Kendall tau C = 0.59, p less than 0.00005). Compared to pure tone audiometry, the threshold had been estimated correctly by ABR in 81% of the patients (+/- 20 dB HL), in 2% the threshold had been estimated to be more than 20 dB better (false negative), while 17% had the threshold estimated more than 20 dB worse (false positive). The median difference between the thresholds obtained was -10 dB HL, the 25% percentile -20 dB HL and the 75% percentile 0 dB HL. In 28 children, behavioural audiometry and pure tone audiometry could be analyzed. 21% were false negative and 50% false positive. The result of the study indicates that ABR is a reliable testing procedure in the estimation of hearing thresholds in children who cannot cooperate sufficiently for pure tone audiometry. ABR gives a more valid estimation of the hearing than behavioural audiometry and is therefore a most valuable testing procedure in paediatric audiology.