Newborn hearing screening by transient evoked otoacoustic emissions: analysis of response as a function of risk factors.

Hearing loss can be considered as the most common birth defect. Early detection of hearing loss by screening at, or shortly after, birth and appropriate intervention are critical to speech, language and cognitive development. In the present study, the characteristics of Transient Evoked Otoacoustic Emissions have been evaluated as a function of known pre- and perinatal risk factors for hearing loss. All newborns were screened for hearing loss using a physiologic test of hearing function, the Transient Evoked Otoacoustic Emissions. A total of 532 consecutive newborn infants received binaural Transient Evoked Otoacoustic Emission testing (262 males, 270 females; mean gestational age 39.2 +/- 2.1 weeks, range 26-43; birth weight: 3,240 +/- 550 g, range 910-4,780). The population examined comprised 448 control infants and 84 high-risk for hearing loss infants (Joint Committee on Infant Hearing 1994 criteria). All Transient Evoked Otoacoustic Emission recordings were performed at comparable postconceptional ages. Audiological screening by Transient Evoked Otoacoustic Emission recording showed an overall 100% sensitivity, 99.02% specificity, with negative and positive predictive values of 100% and 62.5%, respectively. As compared to controls, high-risk infants showed: 1. increased rates of Fail-1 (Transient Evoked Otoacoustic Emissions absent at first examination, 21.4% vs 9.8%, p = 0.004), Fail-2 (Transient Evoked Otoacoustic Emissions absent on retesting: 8.64% vs 1.37%, p = 0.0014), false positives (Transient Evoked Otoacoustic Emissions absent/V wave present: 3.7% vs 0.46%, p = 0.029) and true positives (Transient Evoked Otoacoustic Emissions absent, V wave absent: 2.47% or 24.5 per 1,000 live births vs 0.22% or 2.2 per 1,000 live births, p = 0.013); 2. significantly reduced Transient Evoked Otoacoustic Emission intensity in the 0.7-1 kHz (right side) and 1-2 kHz (left side) frequency ranges. Multivariate logistic regression analysis showed a significant positive correlation between congenital hearing loss and the following risk factors: assisted ventilation lasting > 10 days (Odds ratio 14.8; 95% confidence interval, 4.5-48.8, p < 0.000001), severe birth asphyxia (Odds ratio 5.8; 95% confidence interval; 2.1-16.1; p = 0.0006) and administration of ototoxic drugs (Odds ratio 4.5; 95% confidence interval; 1.4-13.9; p = 0.009). Results of this study confirm the feasibility and accuracy of universal neonatal hearing screening based on recording Transient Evoked Otoacoustic Emissions. These data stress the importance of the risk factors for hearing loss, including prolonged assisted ventilation, ototoxic drugs, and severe birth asphyxia.