Accurate assessment of speech audibility with a hearing aid is one of the most important steps in providing amplification for children with hearing loss. Making the speech spectrum audible across a wide range of listening situations supports positive developmental outcomes.
The most effective methods for estimating the audibility of speech are the use of a probe microphone to measure the response of the hearing aid to speech or a speechlike stimulus in the child's ear, or the use of a real-ear-to-coupler difference (RECD) to apply to hearing aid measurements taken with a speech stimulus in the coupler. By comparing these electroacoustic data with the child's behavioral audiogram, we can optimize the gain and output of the fitting to give the child the best access to speech. An experienced clinician can complete these measures in less than 10 minutes.
Despite the effectiveness and simplicity of probe microphone verification, our clinic still receives frequent requests for functional gain measurements from parents, school districts, and other audiologists. Functional gain is based on measuring pure-tone behavioral thresholds with the hearing aids in a sound field and comparing those thresholds with the child's unaided audiogram.
UNREALISTIC AND INVALID
Prior to the availability of probe microphone verification equipment, functional gain was one of the only tools we had to estimate improvements with amplification, and its limitations were well-known. One of the first published accounts of the drawbacks of functional gain was reported in 1980 (Aust J Audiol 1980;2:56). Even in the linear hearing aids used at that time, aided thresholds that were poorer than the patient's unaided thresholds were observed due to the circuit noise of the hearing aid or the presence of regions of normal hearing in the listener.
More recently, several of my colleagues published an example of linear and nonlinear hearing aids that had similar aided audibility for an average speech signal but functional gain results that made the nonlinear hearing aid look poorer than the linear device (HJ Nov. 2002, p. 38; http://bit.ly/FGain). Essentially, widely varying functional gain results can be observed even when an average speech signal is audible.
This is not the only reason to avoid using functional gain to assess hearing aids in children:
1. Functional gain uses an unrealistic stimulus. Pure-tone stimuli do not accurately estimate the response of a hearing aid to a speech stimulus because of differences in amplitude compression and other signal processing systems within the hearing aid (Ear Hear 2005;26:409). If audibility of speech is our goal, a speech stimulus should be used.
2. Functional gain is tested at a level far below where our patients listen in everyday situations. In addition to differences between speech and pure tones, measuring aided thresholds at soft input levels does not approximate the amount of gain that will be observed in realistic listening environments, where the noise floor is much higher. Expansion and amplitude compression will affect signals presented at soft levels.
3. Probe microphone measures are faster. For children of any age, in situ verification or RECD can be done in a few minutes and provide data across a much wider frequency range.
4. Audiometric data from children are often limited. Given the importance of regular monitoring of unaided behavioral thresholds, spending test time in the sound booth gathering information that can be better approximated using objective methods is not efficient.
5. Infants and children with developmental delays are not able to perform functional gain. Functional gain is contingent upon the child participating in behavioral testing.
6. Aided speech recognition testing is more valid. For children who are able to complete behavioral assessment with their hearing aids, aided speech recognition testing provides a more realistic estimate of how the hearing aid processes speech and how the child uses that input to support perception than functional gain does.
7. Functional gain does not inform programming adjustments to optimize audibility. Speech audibility cannot be estimated from functional gain measures. Therefore, clinicians cannot determine if speech is audible and are also unable to make changes to gain and output settings based on functional gain to enhance the audibility of speech, if needed.
8. There is no standard for what amount of functional gain is sufficient, beyond the belief that lower thresholds are better. Attempts to optimize the programming of the hearing aid to improve processing of soft sound and lower the functional gain thresholds could have unintended effects on the amount of gain available for soft (55 dB SPL)- and average (65 dB SPL)-level speech inputs.
9. Hearing aid signal processing can affect functional gain. Both amplitude compression and feedback management systems can influence the amount of gain that is provided for a pure-tone signal. The presence of these features in nearly all digital hearing aids could result in aided pure-tone thresholds that do not accurately reflect aided sensitivity.
10. We should never perform a test we know is invalid simply because it is requested. If functional gain assessment of amplification is requested by a parent, funding agency, or another professional, we have a responsibility to share our knowledge about the limitations of this method and explain why alternatives like probe microphone measures are more valid and appropriate.
ASSESSING OTHER DEVICES
While probe microphones are a validated alternative to functional gain for assessment of air-conduction hearing aids, cochlear implants, auditory osseointegrated devices, and bone-conduction hearing aids cannot be verified using this method.
Importantly, many of these devices use signal processing, including amplitude compression and feedback management, which can affect how pure-tone inputs are processed. The potential effects of these processing strategies on device settings should be considered. Aided speech recognition testing should also be used whenever possible to estimate how the device processes speech at realistic input levels.