CHAPTER TWENTY-SIX
Mode of operation
200
Mode of operation
26.3.4
Specificity
The specificity of an ABR hearing screening instrument quantifies the ability
of the device to identify ears with normal hearing (the true negative rate).
When there is no hearing loss we can assume that an ABR is present.
Specificity is therefore the probability of a ‘Valid ABR’ (pass) result when
an ABR is truly present. In practice ABR screening test specificity is
dependent on both the amplitude of individuals ABR at the electrodes and
the levels of noise occurring during the test. Both vary according to the
individual, electrode connection, position, resting state of the infant and
the environmental noise. The mean amplitudes of ABR from 270 infant
was determined from healthy normally hearing baby’s ABR recorded by
the Otoport OAE+ABR using our short chirp stimulus at a level of 35dBnHL
and a rate of 52 per second. The normal distribution (spread) of ABR
amplitudes was determined from an historical study of 3200 infant ABR
recordings at 30dBHL*, in particular the probability of small ABR amplitudes
from normal ears. The noise levels recorded by the Otoport OAE+ABR
instrument during tests when the baby was quiet, were averaged. The
probability of the Otoport OAE+ABR normal ABR amplitudes, including the
weakest normal ABRs, against fluctuating noise of the average amplitude
in ‘quiet babies’ was determined to be as below:
All Pass Criteria
Pass probability (%)
99.70
CI(Hi) (%)
99.75
CI(Low) (%)
99.59
Specificity
99.79%
Important Notes:
• High EEG, myogenic or powerline noise levels can greatly
reduce specificity.
• Testing with active infants will reduce specificity, depending on
the movement noise.
• Testing with higher stimulus levels will increase specificity
because the response will be larger but will reduce the range
of threshold elevations to which the instrument is sensitive
(according to the stimulus dBnHL used).