3. Results
In this study, Tone Burst and NB CE-Chirp ABR data of 24 adults (31 ears) were analyzed retrospectively. The tests could not be completed in both ears, since 17 individuals could not tolerate repeated ABR measurements for each ears. The mean age of the participants was detected at 30.4 ± 8.6 (min: 20 max: 48). Test data of 18 ears from 17 men and 13 ears from seven women were analyzed. The research included 18 (58.1%) right ears and 13 (41.9%) left ears. The absolute latencies of peak V with TB stimuli were significantly longer than latencies obtained with NB CE Chirp stimuli at 0.5, 1 and 2 kHz at all sound intensity levels (p<0.001) (Figure 1, 2) . There was no difference between TB ABR and NB CE-Chirp ABR latencies at 4 kHz-20 dB nHL (p=0.374) (Table 1) . The amplitudes of peak V detected with NB-CE Chirp were significantly larger than TB amplitudes in the level of 4 kHz 60 dB nHL (p=0.038). However, there was no significant difference was detected at other frequencies for the amplitude in both TB and NB CE-Chirp ABR (p>0.05).
NB CE-Chirp ABR threshold values closer to PTA 0.5 1, 2 kHz thresholds than the TB ABR thresholds (p<0.001, CI=2.0-5.9 dB)(Table 2) (Figure 3). While the thresholds obtained with PTA, TB and NB CE-Chirp stimuli were showing the highest correlation at the 2 kHz, the least correlation between the measurements was found at 4 kHz (respectively ICC= 0.855 and 0.673), (Table 2). The correction factor, which was calculated from the mean difference in the paired t-test, between thresholds obtained from NB CE-Chirp and TB stimuli and BHT determined by PTA is given in Table 3. In the posthoc power analysis, the power for the difference between the threshold values at all frequency and all sound intensities was measured as 90% and above.
There was no significant difference between thresholds obtained from PTA, TB and NB CE-Chirp for right and left ears at all frequencies (p>0.05, paired t-test). While the most consistent (test-retest reliability) threshold value obtained with PTA at 4 kHz frequency with TB ABR in the right ear (r=0.709, p=0.001), it obtained at 2 kHz with NB CE-Chirp ABR in the left ear (r= 0.743, p= 0.004). The estimated thresholds were similarly correlated with PTA in the right and left ears. The mean test time with NB CE-Chirp ABR was calculated as 23.6 ± 3.9 minutes and the mean test time with Tone Burst ABR was measured as 28.2 ± 4.5 minutes. NB CE-Chirp ABR test time was significantly shorter than the TB ABR test time (p=0.011).
4. Discussion 
Ferm et al. (16) compared NB Chirp and Tone pip ABR findings at 1 and 4 kHz and 40 dB nHL of 42 ears of 30 infants in the newborn screening program. As a result of this study, NB CE-Chirp ABR wave V amplitudes were found to be larger than Tone pip ABR wave V amplitudes (p=0.001). In 40 children with normal hearing, the findings of the 0.5, 1, 2, 4 kHz NB CE-Chirp ABR and Tone-Burst ABR were compared at 80, 60, 40, 20 dB nHL by Rodrigues at al.(12). At all levels except 500 Hz 80 dB nHL, NB CE-Chirp ABR wave V amplitudes were found to be greater than Tone-Burst ABR wave V amplitudes (p<0.05).
In our study, only at 4 kHz 60 dB nHL level, NB CE-Chirp ABR wave amplitudes were found to be greater than TB ABR wave amplitudes (p=0.038). Similar to our research results, Megha et al. (25) found no significant difference in any of the frequencies for amplitude parameter in TB and NB CE Chirp ABR. At high levels of chirp stimulus, desynchronization develops as a result of overstimulation on the cochlear basilar membrane. As a result, it is stated that the wave V amplitudes decrease (25-27). This effect can be observed, especially when a high stimulus level is sent to subjects with normal hearing (25, 27).
Studies comparing NB CE-Chirp ABR and Tone Burst ABR in the literature were frequently conducted with infants. For example, Rodrigues et al. (12) found that NB CE Chirp ABR latencies were shorter than Tone Burst ABR latencies in infants at 0.5, 1, 2 kHz. In this study; the mean wave V latency evoked by 500 Hz (60 dB nHL) NB CE-Chirp was found to be 3.57 (SD: 0.70). In this study, longer wave latencies were found compared to our research. Because the value of the peak V latency varies among different age person (28). Our study was carried out with adults.
Megha et al. (25) conducted a study that researched the effects of noise using NB CE Chirp ABR and TB ABR. Similar to our research, in this study; the mean wave V latency evoked by 500 Hz NB CE-Chirp was found to be 2.45 (SD: 0.68) in the control group with normal hearing. Especially at low frequencies, NB CE-Chirp ABR wave V latencies were observed to be very short. For Tone Burst stimulus, the ABR mean latencies increased with decrease in frequency. But NB CE-Chirp stimulus, the ABR means latencies decreased with decrease in frequency. This is explained by eliminating the delay in the cochlear travelling wave by using different transmission times for each of the chirp stimulus frequencies. The low frequencies in the NB CE-Chirp are being presented early than high frequency octave bands. Thus NB-CE Chirp provides maximum stimulation into the cochlea (12, 25).
Ferm et al. (16) reported that the NB CE-Chirp ABR thresholds were found to be lower than the tone pip thresholds. For this reason, it was emphasized that NB CE-Chirp stimulus can provide closer responses to behavioral thresholds.
In our study, ABR thresholds with Narrow Band CE-Chirp stimulus at 0.5, 1 and 2 kHz were found significantly closer to the behavioral hearing thresholds compared to Tone Burst ABR thresholds (p<0.001, CI=2.0-5.9). Talaat et al. (29) investigated which of the NB CE-Chirp ABR and Tone Burst ABR thresholds were closer to the frequency specific behavioral hearing thresholds. As a result, it was stated that NB CE Chirp-ABR provides higher sensitivity and accuracy than TB ABR in estimating behavioral hearing thresholds in young children. Van et al. (30) compared the proximity of NB CE-Chirp ABR and Tone Burst ABR thresholds to behavioral hearing thresholds in 23 adults with normal hearing. In this study, it was stated that using NB CE-Chirp is more reliable than TB ABR in estimating behavioral hearing thresholds. Similarly, the ASSR thresholds determined by the NB CE-Chirp stimulus are highly correlated with the behavioral thresholds (0.5, 1, 2, 4 kHz) defined with PTA according to the TB ABR thresholds (31, 32).
The hearing thresholds specified as dB nHL in ABR is not equal to behavioral thresholds determined as dB HL. As a result, it is suggested to use the correction factor when calculating BHT thresholds (dB HL) from the thresholds (nHL) determined by ABR (25, 28). The correction factor, which is used to estimate behavioral hearing thresholds with using of NB CE-Chirp ABR, can be reduced by up to 5 dB (21, 16).
The shortening of the test time with NB CE-Chirp stimulus is another advantage. ABR amplitudes are not detectable in 30 % of the click stimulus and need more test time. Click stimulus produces larger and easily detectable amplitudes. Likewise, the mean test time of NB CE-Chirp ABR was detected shorter than the mean test time of TB ABR in our research (p=0.011). Similarly to this study, Zirn et al. (33) compared the duration of the test conducted with both ABR methods in their study including 253 children, and as a result, NB CE-Chirp ABR was found to shorten the recording time. Ferm et al. (16) reported that the use of NB CE-Chirp stimulation produced larger response amplitudes, and therefore increase the signal to noise ratio according to TB clicks, so shortens the test time.