102 Speech Audiometry
In many animals the sense of hearing is adapted for a specific purpose. In the human, the ear is specifically tuned to the speech frequencies (500–4,000 Hz). One of the main functions of the human ear is therefore the perception of speech. Indeed, most of the handicap of hearing loss is due to loss of the ability to perceive the spoken word. Speech audiometry provides a measure of this ability and any corresponding deficit. Voice tests can be considered as a very basic form of speech audiometry. However, in general, speech audiometry implies the formal qualitative assessment of a subject’s perception of speech. It serves as a measure of disability produced by any hearing impairment, but it is not without its limitations, for example, usually tested in a soundproof room and often through earphones delivering the stimulus to each ear separately, which is different to a real-world situation. It is, however, useful in a variety of contexts including the following:
a. Assessment and diagnosis of peripheral and central hearing disorders.
b. Prediction of the usefulness of a hearing aid and as a counselling tool.
c. Evaluation of the benefits of certain surgery (pre- and post-operative assessment), such as stapedectomy and cochlear implantation.
d. Medicolegal assessment and as a means to crosscheck the validity of pure-tone audiometry.
102.1 Materials
1. Instruments Testing is performed in a soundproof room, with the volume controlled from the audiometer, which presents speech material to the subject via loudspeakers or headphones. Live speech material presented by a tester using a microphone is prone to variation in both intensity and accent. Standardised, prepared speech material presented by CD or computer and controlled by the audiometer is much more desirable. The use of headphones, as opposed to the free-field situation with loudspeakers, allows each ear to be tested individually and allows the non-test ear to be masked.
2. Speech material Phonemes are the building blocks of speech and represent the smallest unit of recognisable speech sound (e.g. ay, aw, ah, etc.). There are 49 phonemes in the English language. Speech material is chosen to provide a representative balance of phonemes and can be presented as words, sentences or synthetic sentences which have no meaning. A great number of lists of appropriate speech material have been developed by various agencies, for example, Medical Research Council, Institute of Hearing Research, Fry, Boothroyd and Manchester junior word lists, and Bench, Koval and Bamford (BKB) sentences (University of Manchester). There are also international word lists and ideally word lists in the patient’s first language will provide optimum results.
102.2 Procedure
The subject is seated in a soundproof room and instructed in the test procedure. The recorded word lists are presented to the patient monaurally over headphones at various sound intensities. As speech audiometry is a suprathreshold test, masking of the non-test ear is usually required. Masking sounds in speech audiometry are chosen to try and recreate an appropriate noise background such as speech, cocktail party and babble noise. Pink noise and wide band noise (equal energy for each octave over the hearing range) are usually used when these are unavailable.
The first presentation is usually at 20 to 30 dB greater than the pure-tone average for the frequencies 500, 1,000 and 2,000 Hz. Subsequent presentations are usually made at +10, +20, −10 and −20 dB from this level, although more may be required. The patient is asked to repeat the words as accurately as possible and the percentage of words or phonemes, which are correctly repeated at each sound intensity (dB) is calculated and plotted on a graph.
The graphical display is compared with the calibration graph for that particular machine, which will have been obtained by testing otologically normal individuals on that machine, using the same tapes and test environment. The recorded data can then be used to formulate certain scores, and the shape of the graph used to give information regarding the type of deafness.
