Dr. McCreery is an associate director of audiology and staff scientist at Boys Town National Research Hospital in Omaha, NE.
Over the past few decades, otoacoustic emission (OAE) assessment has become a ubiquitous clinical test of cochlear outer hair cell function in infants and children. Many birthing hospitals incorporate otoacoustic emissions as part of their universal newborn hearing screening program, and most clinicians are comfortable using and interpreting OAE in a pediatric test battery, even if they rarely see children in their practice.
Figure. iStockphoto...Image Tools
One of the main reasons for the widespread use and acceptance of otoacoustic emission assessment is the simplicity of interpretation, with most OAE equipment reporting either a “pass” or “refer” result at the end of the test. Unfortunately, this simplicity can cause us to disengage our critical-thinking skills, leading to errors in how we diagnose hearing loss.
Two primary methods of eliciting otoacoustic emissions are used clinically: distortion-product (DPOAE) and transient-evoked (TEOAE). In DPOAE, two tones are presented at different levels and frequencies. The relationship between their frequencies is selected to elicit a response in the cochlea at a third frequency, where the DPOAE occurs. Different combinations of frequencies prompt responses from different frequency regions of the cochlea.
For TEOAE, clicks or other brief stimuli are used to elicit OAE across a broad frequency range. Whether a clinic uses DPOAE or TEOAE usually depends on the available equipment, as both methods have good evidence to support their use.
If you ask a clinician, most consider an OAE response present if it is 6 dB above the noise floor of the recording, and absent or noisy if the response is less than 6 dB above the noise floor. Sounds simple enough, right?
Figure. Ryan McCreer...Image Tools
Although we recognize an absent OAE as indicating hearing loss and a present OAE as indicating normal hearing, work from Michael Gorga, PhD, and colleagues highlights the fact that the distribution of OAE responses for ears with normal hearing and ears with hearing loss overlaps (Ear Hear 1997;18:440-455 ). In practical terms, this means that there are patients with permanent hearing loss who have otoacoustic emissions meeting the 6-dB single-to-noise ratio (SNR) criterion, and patients with normal hearing who do not. How can clinicians tell the difference?
The most obvious answer to this question is that OAE should probably never be the only assessment tool used in a diagnostic setting. Discussion of the cross-check principle, which says never to base a diagnosis on a single clinical test, has been going on for 40 years.
While some may think that the suggestion to use multiple tests is overly simplistic, our clinic has received reports over many years where hearing loss was diagnosed or excluded based on OAE alone. In infants, immittance testing and some method of estimating threshold via auditory brainstem response should be part of the diagnostic test battery. In older children, a combination of immittance and behavioral testing should be used.
DPOAE CLINICAL FORM
Another solution to improve our interpretation of OAE results is by applying research data. Dr. Gorga and colleagues’ work reports the distributions of OAE responses at different frequencies for people with normal hearing and people with hearing loss. Using this data, our clinicians developed a form to plot OAE responses that are at least 6 dB above the noise floor (see image on page 16). The overlap in responses between ears with normal hearing and ears with hearing loss is noted by the area marked “B.”
Figure. The B area r...Image Tools
OAE responses that have a level (dB) above the marked area are considered to be within the normal range; responses that fall below that area are considered abnormal. Responses within the “B” area cannot be determined as normal or abnormal because that area represents the levels where both normal and abnormal groups overlap.
As an example, consider “Patient A” and “Patient B” who had OAE responses measured in your clinic. At 4,000 Hz, Patient A has an OAE level of 12 dB and a noise floor of 3 dB for a 9-dB SNR. Patient B has an OAE level of -10 dB and a noise floor of -19 dB for the same 9-dB SNR.
Using only the 6-dB criterion, both patients would be considered to have present OAE responses at 4,000 Hz. If those results are plotted on the OAE form, Patient A has a response that falls within the normal range (12 dB) and Patient B has a response that falls within the range for people with hearing loss. Using this approach to interpret OAE data can help improve our classification of hearing loss versus normal hearing.
The clinical otoacoustic emission form has important limitations that should be considered as well. First, the data used to develop the form were DPOAE data, so TEOAE data should not be plotted on the form. TEOAE data that could be used to develop a similar form is available in a paper from Debra Hussain and colleagues (Ear Hear 1998;19:434-449 ).
Responses without at least a 6-dB SNR should not be plotted on the form, including high OAE levels with a high noise floor and very low OAE levels near the noise floor of the OAE equipment.
Using the cross-check principle and the DPOAE clinical form can help improve our confidence in and interpretation of clinical test data.
Read past Building Blocks columns in a special collection on the HJ website: bit.ly/HJBuildingBlocks.
To obtain an electronic copy of the DPOAE form for your clinic, e-mail Nancy.Johnson@boystown.org.
© 2013 by Lippincott Williams & Wilkins, Inc.