Speech audiometry.

Speech audiometry--an essential part of the battery of comprehensive audiometric tests--evaluates speech reception, recognition, and understanding. Speech audiometry results and pure-tone thresholds help otologists determine the level of hearing dysfunction and select the appropriate hearing aids for individual patients. The role of speech processing and acoustics in determining the degree of hearing loss is a complex subject that is beyond the scope of this article. Instead, our focus is on the most commonly used speech tests in clinical practice: the speech reception threshold (SRT) of spondee words, the word recognition score (WRS) of phonetically balanced words, most comfortable loudness (MCL), and uncomfortable loudness (UCL).

The SRT determines levels of MCL and UCL. Other tests for assessing speech understanding in noise, for fitting hearing aids, and for evaluating cochlear implant patients are the hearing-in-noise test, (1) the speech-perception-in-noise test, (2) and the sentence-recognition score. (3)

The American Speech and Hearing Association's recommendations for the WRS require the use of standard speech lists (the most common are the NU-6 and CID W-22 lists) that are delivered to earphones, insert phones, or the sound field by audiotapes or compact disks rather than by live voices to ensure reliability, consistency, and accuracy. Significant variability in compliance with these recommendations by audiologists has been reported. (4)

Patients commonly respond in an open-set format; a closed-set format can be used for children and geriatric patients (especially those with memory difficulties) and in special situations when an open-set format is not applicable.

The SRT is normally 10 dB higher than the pure-tone average (PTA) at 0.5, 1, and 2 kHz of the corresponding audiogram. A difference of more than 10 dB raises questions about test reliability and pseudohypoacusis.

The WRS is determined at the 40-dB sensation level (dBSL) with masking of the nontest ear when necessary. The dBSL should be below the UCL level, and the MCL level should be used to determine WRS, even if it is less than 40 dB.

[FIGURE OMITTED]

The WRS is classified as normal (>90%),fair to good (50 to 80%), and poor (<50%). A drop in WRS of more than 20% at higher presentation levels (rollover test) or with competing conversational background noise is suggestive of retrocochlear dysfunction. A 20% difference between two tests is statistically significant when using 50-word lists .5 A disproportionate difference between pure-tone thresholds/averages and WRS raises questions about cochlear (immune-mediated) hearing loss, (6) retrocochlear dysfunction, and pseudohypoacusis.

Testing speech recognition in noise is more clinically utilized and will be the subject of a future clinic. A cursory method of assessing speech recognition in noise is to deliver a set of common sentences about activities of daily living with a live voice in the presence of background conversational noise (signal-to-noise level: 10 dB) and qualitatively evaluate the response. This is helpful in guiding hearing rehabilitation, especially for patients with severe high-frequency sensorineural hearing loss and a WRS in the 50 to 70% range. It is important to understand that hearing aids do not improve WRSs, so patients must be counseled against buying new, "more powerful" digital hearing aids. Such a purchase will only result in needless frustration and cost.

A loss of speech discrimination, irrespective of pure-tone thresholds, should be treated as an urgent otologic condition that requires a medical evaluation and the use of oral and/or intratympanic steroids. (6) If this should fail, the patient becomes a candidate for cochlear implantation.

References

(1.) Nilsson M, Soli SD, Sullivan JA. Development of the Hearing in Noise Test for the measurement of speech reception thresholds in quiet and in noise. J Acoust Soc Am 1994;95:1085-99.

(2.) Bilger RC, Nuetzel JM, Rabinowitz WM, Rzeczkowski C. Standardization of a test of speech perception in noise. J Speech Hear Res 1984;27:32-48.

(3.) Speaks C. Intelligibility of filtered synthetic sentences. J Speech Hear Res 1967;10:289-98.

(4.) Martin FN, Champlin CA, Chambers JA. Seventh survey of audiometric practices in the United States. J Am Acad Audiol 1998;9:95-104.

(5.) Thornton A, Raffin MJ. Comment on "A quantitative examination of the sources of speech discrimination test score variability." Ear Hear 1982;3:340.

(6.) Hamid MA. Recent advances in medical otology-neurotology. Mediterranean Journal of Otology 2005; 1:45-53.

Mohamed A. Hamid, MD, PhD; Kenneth H. Brookler, MD