Optimizing Clinical Interpretation of Distortion Product Otoacoustic Emissions in Infants.

Most included ears had normal hearing, and only 85 ears had hearing loss, so the hearing-loss classification results depend on a small loss sample. Most hearing-loss ears had conductive loss, with 67 conductive, 10 sensorineural, and 8 mixed ears, so the findings for sensorineural or mixed loss rest on small numbers. The study could not evaluate test performance for normal hearing versus pure sensorineural loss because only 10 ears had pure sensorineural loss. Of 975 enrolled ears, 456 were analyzed; exclusions included 18 ears without DPOAE data, 17 without ABR data, 2 with auditory neuropathy spectrum disorder, 5 tested outside the first visit age range, 473 not meeting normal hearing criteria, and 4 hearing-impaired ears missing later test data. Hearing status was assigned from an overall clinical diagnosis that combined first-visit tests and later confirmation, and some auditory brainstem response and behavioral hearing test thresholds did not agree. In about 80% of cases with conductive loss diagnosed by auditory brainstem response, hearing was normal on later behavioral testing, so conductive status at one visit may not reflect stable hearing. Absorbance was tested only at ambient pressure, not at tympanometric peak pressure. Auditory brainstem response thresholds were not always measured at all frequencies because testing depended on infant state and time. Later diagnostic accuracy analyses were not done because too few ears had hearing loss at 9 to 12 months.

The supplied text lists research support from the American Recovery and Reinvestment Act and NIH extramural funding. It does not provide a commercial conflict of interest statement or author disclosure list.

Do not treat the 100% sensitivity for sensorineural or mixed loss as proof that DPOAEs detect all infant hearing loss. That result excluded pure conductive ears, and the sensorineural or mixed group contained 10 ears with sensorineural loss and 8 ears with mixed loss.