The Relationship Between Pure Tone Thresholds and Speech Measurements
PT thresholds are often compared with speech audiomet- ric test results. The two most common comparisons are with speech reception thresholds (SRT) and suprathreshold word-recognition scores (WRSs). (See Chapter 5 for a com-prehensive review of speech audiometry.)
SRTs obtained using spondaic words (or spondees) agree well with PT thresholds for low frequencies. Spond- ees are easily recognized; listeners only need to recognize the vowels to identify these words correctly. Because of the importance of the vowels at low intensities, spondee thresh- olds are found to agree closely with the average of PT thresh- olds for 500 and 1,000 Hz (Carhart and Porter, 1971). In the event of a rising audiogram, better agreement between the spondee and PT thresholds is the average for 1,000 and 2,000 Hz. Spondee thresholds and a two-frequency PTA, as noted earlier, nearly always agree within ±10 dB in coop- erative examinees. This agreement makes the threshold for spondaic words an excellent check on the validity and reliability of the audiogram. This comparison is important for most children. It is also a valuable tool for assessing the reliability of PT thresholds in adults who demonstrate inconsistent puretone responses or who may present with pseudohypacusis (Schlauch et al., 1996).
Suprathreshold word-recognition performance is assessed in most clinical settings by scoring a client’s ability to repeat back a list of monosyllabic words. WRSs provide a valid estimate of speech understanding ability (Wilson and Margolis, 1983) and quantification of the distortion, if any, caused by sensory/neural hearing loss. WRSs are corre- lated with puretone audiometric thresholds in persons with cochlear losses (Pavlovic et al., 1986), but individuals’ scores vary considerably depending on the type of damage to the auditory system. If the words are presented at a level high enough to make the speech sounds audible (overcoming the attenuation caused by the loss), persons with mild cochlear hearing loss are expected to have high WRSs, and those with severe to profound losses are likely to have fairly low scores. Dubno et al. (1995) and Yellin et al. (1989) have published tables relating WRSs and the average of PT thresholds for 500, 1,000, and 2,000 Hz for groups of persons with typical cochlear losses. WRSs that are abnormally low for a given PTA are associated with a variety of conditions including an acoustic tumor, multiple sclerosis, Ménière’s disease, auditory neuropathy, and cochlear dead regions (Moore, 2004), to name a few. When there are dead regions (areas of missing inner hair cells in the cochlea), PT thresholds may appear artificially better than expected because of the spread of energy along the cochlea. Healthier cochlear cells adjacent to the missing cells will elicit a response to pure- tones presented at the dead region frequency.
SRTs obtained using spondaic words (or spondees) agree well with PT thresholds for low frequencies. Spond- ees are easily recognized; listeners only need to recognize the vowels to identify these words correctly. Because of the importance of the vowels at low intensities, spondee thresh- olds are found to agree closely with the average of PT thresh- olds for 500 and 1,000 Hz (Carhart and Porter, 1971). In the event of a rising audiogram, better agreement between the spondee and PT thresholds is the average for 1,000 and 2,000 Hz. Spondee thresholds and a two-frequency PTA, as noted earlier, nearly always agree within ±10 dB in coop- erative examinees. This agreement makes the threshold for spondaic words an excellent check on the validity and reliability of the audiogram. This comparison is important for most children. It is also a valuable tool for assessing the reliability of PT thresholds in adults who demonstrate inconsistent puretone responses or who may present with pseudohypacusis (Schlauch et al., 1996).
Suprathreshold word-recognition performance is assessed in most clinical settings by scoring a client’s ability to repeat back a list of monosyllabic words. WRSs provide a valid estimate of speech understanding ability (Wilson and Margolis, 1983) and quantification of the distortion, if any, caused by sensory/neural hearing loss. WRSs are corre- lated with puretone audiometric thresholds in persons with cochlear losses (Pavlovic et al., 1986), but individuals’ scores vary considerably depending on the type of damage to the auditory system. If the words are presented at a level high enough to make the speech sounds audible (overcoming the attenuation caused by the loss), persons with mild cochlear hearing loss are expected to have high WRSs, and those with severe to profound losses are likely to have fairly low scores. Dubno et al. (1995) and Yellin et al. (1989) have published tables relating WRSs and the average of PT thresholds for 500, 1,000, and 2,000 Hz for groups of persons with typical cochlear losses. WRSs that are abnormally low for a given PTA are associated with a variety of conditions including an acoustic tumor, multiple sclerosis, Ménière’s disease, auditory neuropathy, and cochlear dead regions (Moore, 2004), to name a few. When there are dead regions (areas of missing inner hair cells in the cochlea), PT thresholds may appear artificially better than expected because of the spread of energy along the cochlea. Healthier cochlear cells adjacent to the missing cells will elicit a response to pure- tones presented at the dead region frequency.