Rapid NHS using AABR with high stimulus repetition rates and automated signal detection

The automated auditory brainstem response (AABR) is an important tool for detecting hearing impairment in newborns, but it is often hindered by unacceptably long test times. On average, it takes 3 to 5 minutes per infant to acquire sufficient data to determine if an AABR waveform is present or absent. The successful reduction of this average test time would be of significant benefit to both the institutions and the societies providing UNHS services. Our aim was to determine if increased stimulus repetition rates and automated signal detection could be used to reduce AABR test time. To do this, we conveniently sampled 80 neonates (aged 0 to 3 days) from the Mater Mothers’ Hospital in Brisbane, Australia, all of whom had passed a screening test in both ears using that hospital’s hearing screening device. We reassessed these neonates in one ear (right or left) using our purpose built AABR device, obtaining AABR waveforms from each neonate using 35 dBnHL conventional click stimuli at 33 and 90 clicks per second (cps) and 35 dBnHL maximum length sequence (MLS) click stimuli at 180, 250, 500 and 833 cps with both bipolar and unipolar reconstructions. We then analysed these AABR waveforms off-line using subjective detection by a human observer, and the automatic signal detection techniques of Fsp and cross-correlation, to determine which combination of these stimulus repetition rates and automatic detection techniques provided the shortest AABR test time. We will present the results of our study, its implications for the use of the AABR in UNHS, and our plans for future research