Abstract
The auditory function of 370 infants, drawn from both low and high risk groups, was tested before postnatal discharge using three tests: standard auditory brain stem responses (ABR), automated analysis of ABR, and automated analysis of evoked otoacoustic emissions (OAE). All infants failing any neonatal test had further audiological evaluation. Follow up information was also available on those who passed neonatal tests. Automated OAE testing of both ears was quickest (median 12.5 minutes) and least invasive (no scalp electrodes). Bilateral failure rates (and upper 95% confidence limits) with a stimulus 35-36 dB above normal hearing threshold level (nHL) were 3.0% (4.6) with automated OAE, 3.2% (5.1) with ABR, and 2.7% (4.4) with automated ABR. Automated OAE was the test most sensitive for subsequently confirmed hearing impairment. Sequential testing with automated OAE followed, in those failing this test, by automated ABR would have provided a screening test for substantial hearing impairment with a specificity greater than 99% in this population. Possible application as a universal screen is discussed.