OBJECTIVE: The aim of this study was to assess the diagnostic value of click-evoked otoacoustic emissions (CEOAEs) and tone-burst evoked otoacoustic emissions (TBOAEs) in identifying the residual hearing of subjects with partial deafness (PD)--a condition in which subjects have normal thresholds at low frequencies and severe-to-profound sensorineural hearing loss at high frequencies. DESIGN: Otoacoustic emissions (OAEs) were measured in 23 subjects with severe-to-profound sensorineural hearing loss, 46 with PD, and 15 with normal hearing (NH). The PD subjects were divided into three subgroups according to the frequency at which hearing loss started: PD250 (NH up to 250 Hz)--20 ears; PD500 (NH to 500 Hz)--18 ears; and PD1000 (NH up to 1000 Hz)--20 ears. Standard-click stimuli, and 0.5- and 1-kHz tone bursts (average amplitude 80 ± 3 peak dB SPL, nonlinear averaging protocol), were used. The tone bursts were four cycles long with equal rise/fall times and no plateau. Recordings were performed in two acquisition windows: a standard one 20-msec wide for clicks and 1-kHz tone bursts, and one 30-msec wide for 0.5-kHz tone bursts. OAE response levels, signal-to-noise ratios, and reproducibility were examined in terms of wide-band responses and in terms of half-octave bands centered at 0.5 and 1 kHz. Receiver operator characteristic analysis was used to determine which type of stimuli best differentiates partially deaf subjects from subjects with severe-to-profound sensorineural hearing loss through the range 125 to 8000 Hz. RESULTS: Nearly all recordings from groups PD500 and PD1000 showed 0.5-kHz TBOAEs. By contrast, 1-kHz TBOAEs and CEOAEs were generally found only in the PD1000 group. It was also possible to detect 0.5-kHz TBOAE responses in approximately 50% of ears from the PD250 group. Receiver operator characteristic analysis demonstrated that click and 1-kHz tone bursts provide a good diagnostic measure of residual hearing even when hearing loss starts as low as 1 kHz; moreover, the 0.5-kHz TBOAE could identify residual hearing when hearing loss started just > 0.5 kHz, a situation in which clicks failed to elicit a response. In the case of partially deaf subjects, diagnosis was more accurate when OAEs were analyzed by 1/2 octave bands. Furthermore, the use of a 0.5-kHz tone burst gave responses in the highest number of subjects, even when there were hearing losses in neighboring bands. CONCLUSIONS: The results of this study indicate that a 0.5-kHz TBOAE is a more powerful test than the standard CEOAE when cochlear function at low frequencies is of interest. The 0.5-kHz TBOAE may be used to identify partial deafness in patients who generally fail to show a response to the commonly used clicks. In addition, use of 1/2 octave-band filtering can increase the reproducibility and power of the test.
OBJECTIVE: The aim of this study was to assess the diagnostic value of click-evoked otoacoustic emissions (CEOAEs) and tone-burst evoked otoacoustic emissions (TBOAEs) in identifying the residual hearing of subjects with partial deafness (PD)--a condition in which subjects have normal thresholds at low frequencies and severe-to-profound sensorineural hearing loss at high frequencies. DESIGN: Otoacoustic emissions (OAEs) were measured in 23 subjects with severe-to-profound sensorineural hearing loss, 46 with PD, and 15 with normal hearing (NH). The PD subjects were divided into three subgroups according to the frequency at which hearing loss started: PD250 (NH up to 250 Hz)--20 ears; PD500 (NH to 500 Hz)--18 ears; and PD1000 (NH up to 1000 Hz)--20 ears. Standard-click stimuli, and 0.5- and 1-kHz tone bursts (average amplitude 80 ± 3 peak dB SPL, nonlinear averaging protocol), were used. The tone bursts were four cycles long with equal rise/fall times and no plateau. Recordings were performed in two acquisition windows: a standard one 20-msec wide for clicks and 1-kHz tone bursts, and one 30-msec wide for 0.5-kHz tone bursts. OAE response levels, signal-to-noise ratios, and reproducibility were examined in terms of wide-band responses and in terms of half-octave bands centered at 0.5 and 1 kHz. Receiver operator characteristic analysis was used to determine which type of stimuli best differentiates partially deaf subjects from subjects with severe-to-profound sensorineural hearing loss through the range 125 to 8000 Hz. RESULTS: Nearly all recordings from groups PD500 and PD1000 showed 0.5-kHz TBOAEs. By contrast, 1-kHz TBOAEs and CEOAEs were generally found only in the PD1000 group. It was also possible to detect 0.5-kHz TBOAE responses in approximately 50% of ears from the PD250 group. Receiver operator characteristic analysis demonstrated that click and 1-kHz tone bursts provide a good diagnostic measure of residual hearing even when hearing loss starts as low as 1 kHz; moreover, the 0.5-kHz TBOAE could identify residual hearing when hearing loss started just > 0.5 kHz, a situation in which clicks failed to elicit a response. In the case of partially deaf subjects, diagnosis was more accurate when OAEs were analyzed by 1/2 octave bands. Furthermore, the use of a 0.5-kHz tone burst gave responses in the highest number of subjects, even when there were hearing losses in neighboring bands. CONCLUSIONS: The results of this study indicate that a 0.5-kHz TBOAE is a more powerful test than the standard CEOAE when cochlear function at low frequencies is of interest. The 0.5-kHz TBOAE may be used to identify partial deafness in patients who generally fail to show a response to the commonly used clicks. In addition, use of 1/2 octave-band filtering can increase the reproducibility and power of the test.