Tehran University Medical Journal (Jul 2007)
Accuracy of Cortical Evoked Response Audiometry in estimating normal hearing thresholds
Abstract
Background: Cortical Evoked Response Audiometry (CERA) refers to prediction of behavioral pure-tone thresholds (500-4000 Hz) obtained by recording the N1-P2 complex of auditory long latency responses. CERA is the preferred method for frequency–specific estimation of audiogram in conscious adults and older children. CERA has an increased accuracy of determination of the hearing thresholds of alert patients with elevated hearing thresholds with sensory hearing loss; however few publications report studies regarding the use of CERA for estimating normal hearing thresholds. The purpose of this research was to further study the accuracy of CERA in predicting hearing thresholds when there is no hearing loss. Methods: Behavioral hearing thresholds of 40 alert normal hearing young adult male (40 ears) screened at 20 dB HL in 500-8000Hz, predicted by recording N1-P2 complex of auditory evoked long latency responses to 10-30-10 ms tone bursts. After CERA, pure tone audiometry performed by other audiologist. All judgments about presence of responses performed visually. Stimulus rate variation and temporary interruption of stimulus presentation was used for preventing amplitude reduction of the responses. 200-250 responses were averaged near threshold. Results: In 95% of the hearing threshold predictions, N1-P2 thresholds were within 0-15 dB SL of true hearing thresholds. In the other 5%, the difference between the CERA threshold and true hearing threshold was 20-25 dB. The mean threshold obtained for tone bursts of 0.5, 1, 2 and 4 kHz were 12.6 ± 4.5, 10.9 ± 5.8, 10.8 ± 6.5 and 11.2 ± 4.1 dB, respectively, above the mean behavioral hearing thresholds for air-conducted pure tone stimuli. Conclusion: On average, CERA has a relatively high accuracy for the prediction of normal hearing sensitivity, comparable to that of previous studies performed on CERA in hearing-impaired populations.
