[Early detection of pediatric hearing loss. Visual and automated procedures compared]

Abstract

The recording of otoacoustic emissions is suitable for the detection of hearing loss in small children. The test meets the following requirements for hearing screening: it is carried out in a few minutes, specialized personnel is not necessary, the results do not depend on the vigilance of the child, and total costs are comparably low. However, the choice of a suitable device is quite essential. A test of the ILO88, ILO92, Echosensor (Otodynamics Ltd.), and Echoscreen (Madsen) devices was performed with 102 children (aged < 1 year). Additionally, a software package (Otoclass) for offline analysis of ILO88 data files was tested. The results indicate that all devices applying OAE as a screening measure was able to detect every ear (n = 25) with a hearing loss indicated by the outcome of a control BERA (stimulus level 35 dB nHL), thus reaching a sensitivity equal to 100%. The specificity of the different OAE devices depends on the underlying detection strategy. The best results were achieved with the automated Echoscreen device (95.9%) followed by the Otoclass analysis software (94.2%). The Echosensor device failed in our study to provide good specificity (77.3%). Reflex audiometry, which is favored by pediatricians in Germany, when used alone is completely inadequate as a screening method, even if conducted by a well-trained investigator. In our study, only 61.5% of the children with hearing loss were detected with reflex audiometry, and 42.7% of the children with normal hearing were misclassified. These results deviated from the results presented in ref. 13 to a large extent, as the Hanover group attested sensitivity and specificity of 100% for reflex audiometry (HNO 43, Reuter et al.). The deviating results are discussed in detail.