Chapter 45. Audiologic Testing

Otolaryngologists often rely on audiologic test results to determine the course of treatment for a given patient. Many of the tests constituting the diagnostic audiologic battery of 20 years ago have now been replaced with newer procedures with greater specificity, sensitivity, and site of lesion accuracy. This is exemplified by the fact that the terms “sensory” or “neural” can now frequently replace the term “sensorineural.” In addition, audiologic tests have gone beyond the realm of identifying anomalies in structure to identifying anomalies in function. The logical extension of this advancement is to provide the audiologist and otolaryngologist with information related to prognosis and rehabilitation.

Audiologic tests can be classified according to measures of hearing threshold, suprathreshold recognition of speech, assessment of middle ear function, assessment of cochlear function, determination of neural synchrony and evaluation of vestibular function. The test correlates associated with these measures are pure-tone audiometry, speech recognition, the immittance battery, otoacoustic emissions, electrophysiology, videonystagmography, and rotary chair assessment. The latter two procedures are discussed elsewhere in this textbook.

Audiologic test results should always be interpreted in the context of a battery of tests because no single test can provide a clear picture of a specific patient. In addition, the combination of objective and subjective (behavioral) tests provides a cross-check of the results. There are no age restrictions for audiologic testing; it is now possible and recommended to test newborns within days of birth.

The audiogram is a graph that depicts threshold as a function of frequency. Threshold is defined as the softest intensity level that a pure tone (single frequency) can be detected 50% of the time. Intensity is designated on a normalized decibel hearing level (HL) scale that takes into account the differences in human sensitivity as a function of frequency. The typical range of frequencies tested does not cover the entire range of human hearing (20–20,000 Hz). Instead, the range includes the frequencies considered to be essential for understanding speech (250–8000 Hz). Most testing is administered at discrete octave frequencies. However, when threshold differences between adjacent octaves exceed 15 dB, inter-octave frequencies should be tested. This is particularly true at 3000 and 6000 Hz, where “notches” in audiometric configuration often typify noise-induced hearing loss. Thresholds are measured clinically in 5-dB steps. There is a test-retest variability of ±5 dB. Therefore, a change of 10 dB may not necessarily represent a true threshold shift.

Thresholds can be obtained using air conduction (AC) or bone conduction (BC). Sound transmission via earphones, foam inserts, or loudspeakers requires the movement of air molecules; therefore, it is termed air conduction. This testing assesses the entire auditory system from the outer ear to the auditory cortex. Testing through loudspeakers (sound field) cannot isolate differences between ears. The advantages of insert earphones over over-the-ear (supra-aural) earphones include the prevention of collapsing ear canals, greater attenuation from ambient noise, and greater interaural attenuation (the loss of sound energy that occurs as the signal travels from one ear to the other either around the head or through the bones of the skull). Interaural attenuation is also referred to as “crossover.” The amount of interaural attenuation varies as a function of the transducer type and frequency; it is typically 0 dB for bone conduction, 40–60 dB for supra-aural earphones, and 55–70 dB for insert earphones. AC thresholds are marked on the audiogram with an “O” for the right ear and an “X” for the left ear. BC thresholds are obtained using a small vibrator placed on the forehead or on the mastoid bone. Thresholds are typically indicated on the audiogram with the symbols “<” and “>” (unmasked) or “[” and “]” (masked). Because the skull vibrates as a whole, BC thresholds primarily reflect the contribution of the inner ear, mostly bypassing the function of the outer and the middle ear. The comparison of AC thresholds and BC thresholds provides an initial differentiation between conductive, mixed, and sensorineural involvement. Sensorineural hearing loss is characterized by equivalent air and bone conduction (ie, air–bone gaps of less than 10 dB). Conductive hearing loss is characterized by BC thresholds within normal limits, with a concurrent gap between the poorer AC and better BC thresholds of 10 dB or more. A mixed hearing loss contains air–bone gaps with the bone conduction thresholds outside of the normal range. Figure 45–1 (A–D) shows audiograms depicting normal hearing and these three types of hearing loss.

Both air and bone conduction thresholds may be obtained using an approach that ascends or descends in intensity but are typically determined using a bracketing technique. If tones are presented at high intensity levels, both air- and bone-conducted stimuli can evoke vibrotactile sensations. For AC, vibrotactile thresholds may occur at 90 dB HL at 250 Hz, and at 110 dB HL at 500 and 1000 Hz. For BC, vibrotactile thresholds may occur at 30–35 dB HL at 250 Hz, 55 dB HL at 500 Hz, and 65–70 dB at 1000 Hz. Therefore, patients with severe hearing loss may appear to respond at lower (softer) levels than their true auditory thresholds. For that reason, the tester should ask the patient whether the stimulus was heard or felt when approaching the aforementioned intensity levels. Furthermore, BC greater than approximately 45–60 dB HL in the lower frequencies and 70–75 dB HL in the mid and higher frequencies cannot be measured due to equipment output limits for bone-conducted stimuli. Thus, listeners with severe or profound losses may have real, but nonmeasurable, air–bone gaps, and one must not automatically assume that a profound hearing loss is exclusively sensorineural. This is one of several reasons why a battery of diagnostic test results should always be considered, as opposed to any single measure.

Masking

One of the most important yet confusing aspects of hearing testing is to ensure that the auditory function of each ear is measured independently. In some situations, a noise is presented to the non-test ear to prevent it from responding to a signal presented to the test ear. This is referred to as masking. Masking is required for AC whenever the difference between the air conduction presentation level and the non-test ear BC thresholds exceeds approximately 40 dB for the lower frequencies and 60 dB for the higher frequencies. For BC testing, masking should be used whenever there is any difference in the AC and BC thresholds, since there is essentially no interaural attenuation by bone conduction. When the masking presented to the non-test ear crosses over to the test ear, a masking dilemma results. Usually, this occurs when a patient has a large bilateral conductive hearing loss. The use of insert earphones greatly minimizes these occurrences because of the greater interaural attenuation they provide. Failure to mask appropriately may have potentially serious medical and audiologic consequences.

Categories of Hearing Loss

Table 45–1 provides a general guideline for interpreting degrees of hearing loss based on audiometric findings. Levels should be categorized somewhat more stringently for children.

One commonly used speech measure is the speech reception threshold (SRT). The SRT is the lowest intensity level at which a patient can correctly repeat 50% of common bisyllabic words such as “hotdog” or “baseball.” These results should correspond with pure tone thresholds. However, caution must be exercised in using the SRT as the only indication of hearing sensitivity. An example of how the SRT can provide misleading information is shown in Figure 45–2. Note that for this patient, the SRT is 5 dB, well within the range of normal hearing. This value reflects the normal auditory sensitivity at 500 and 1000 Hz. Further inspection of the audiogram illustrates that this patient has a moderate hearing loss above 1000 Hz and will have considerable difficulty hearing in many acoustic environments. The only true purpose of the SRT is to validate the pure-tone findings. A variation of the SRT is the speech detection threshold (SDT) or speech awareness threshold (SAT), which is the softest level at which a person detects (as opposed to understands) the presence of speech sounds. Pure-tone testing and SRT, SAT, or SDT yield information about hearing sensitivity. Most listening is actually done at suprathreshold levels. Therefore, to define a person's true auditory capacity, measures reflecting suprathreshold listening and clarity must be included. Word recognition testing (formerly referred to as speech discrimination testing) assesses a patient's ability to identify monosyllabic words. A list of words are typically presented to the patient at approximately 40 dB above the SRT, or at a comfortable listening level if 40 dB is either too loud for the patient or incongruous with the audiometric configuration.

The type and degree of hearing loss can often affect word recognition scores. Given the relatively low face validity of monosyllabic word testing (because we do not typically listen to single-syllable words all day), the audiologist can further define hearing by assessing conversational speech or sentence recognition both in quiet and in noise. This information can provide the most useful prognostic data for the rehabilitation plan.

Nonorganic (Functional) Hearing Loss

Pseudohypacusis is defined as functional hearing loss. Occasionally, patients willfully or subconsciously exaggerate their hearing loss. The signs in test behavior that suggest a functional component include (1) inconsistent responses, (2) significant differences between the thresholds obtained using ascending and descending administration of test stimuli, (3) a discrepancy of more than 8 dB between the SRT and the pure-tone average of 500–2000 Hz, and/or (4) a positive Stenger test. The Stenger test may be used to identify unilateral or asymmetrical functional hearing loss. It is based on the concept that when both ears are stimulated simultaneously by a tone equal in frequency and phase, the auditory percept is lateralized to the ear with better hearing. Systematic manipulation of the relative intensities delivered to each ear provides the audiologist with an estimate of the true threshold in the ear that has a more significant hearing loss. When speech stimuli are used, the test is called a Speech Stenger test or a Modified Stenger test. Other objective measures that may disclose functional involvement include acoustic reflexes, auditory brainstem responses, and otoacoustic emissions. These tests are discussed later in this chapter.

Acoustic immittance testing consists of tympanometry, acoustic reflexes, and static compliance. These tests measure the function of the tympanic membrane, middle ear, and acoustic reflex arc pathway. They are not direct measures of hearing sensitivity.

Tympanometry

Tympanometry is based on the amount of sound reflected back from the tympanic membrane when an 85-dB sound pressure level (SPL) low-frequency (226-Hz) probe tone is introduced into the sealed ear canal and pressure in the ear canal is varied. When the pressure in the ear canal corresponds with the pressure in the middle ear cavity, the tympanic membrane is at its most compliant point and thus absorbs, rather than reflects, the most sound. The tympanometric peak, or maximum flow of acoustic energy into the middle ear, occurs when the pressure in the ear canal and middle ear is equal. If eustachian tube function is normal, peak pressure occurs near 0 daPa. If the middle ear is not properly aerated, the middle ear pressure will be negative (>100 daPa). Thus, the ear canal pressure corresponding to the tympanometric peak provides an estimate of middle ear pressure. For infants and neonates, tympanograms should be obtained using a higher-frequency probe tone (660 or 1000 Hz) due to resonant differences in small ear canals.

Classification

Traditionally, tympanograms have been classified as Type A, B, or C (Figure 45–3). Some clinicians prefer describing the tympanogram in a more specific narrative form.

Type A

Type A tympanograms have normal peak height (reflecting compliance) and pressure. Variations of the Type A tympanogram may be normal in pressure but shallow (AS), reflecting otosclerosis or middle ear effusion, or may be peaked very high (AD), reflecting ossicular discontinuity or a monomeric eardrum.

Type B

The Type B tympanogram is flat in appearance, indicating lack of compliance. The volume measurement that is simultaneously performed with tympanometry helps to differentiate between a flat tympanogram suggesting an intact eardrum with middle ear effusion versus a perforated eardrum or ear with a patent ventilating tube.

Type C

The Type C tympanogram has negative peak pressure suggesting inadequate aeration of the middle ear space.

Acoustic Reflex

An acoustic reflex occurs when the stapedius muscle contracts in reaction to a loud sound. Acoustic reflex thresholds refer to the softest intensity levels that can trigger the response. They usually occur at 70–90 dB above the patient's hearing threshold. When the stapedius muscle contracts, it stiffens the ossicular chain, thus decreasing compliance. The change in compliance coincident with the presentation of an intense acoustic signal is measured with the same instrument used for tympanometry. Because monaural stimulation results in contraction of the muscles in both ears, the reflex can be measured either ipsilaterally or contralaterally. When the reflex is recorded in the stimulated ear, it is called an ipsilateral reflex; if it is recorded in the opposite ear, it is called a contralateral reflex. The following four stimulus-probe configurations can be measured: (1) right ipsilateral (stimulus to the right ear, probe to the right ear); (2) right contralateral (stimulus to the right ear, probe to the left ear); (3) left ipsilateral (stimulus to the left ear, probe to the left ear); and (4) left contralateral (stimulus to the left ear, probe to the right ear). Figure 45–4 depicts the acoustic reflex arc. Knowledge of this pathway allows the clinician to compare the results of the various testing configurations to interpret the findings.

Patients with mild or moderate cochlear (sensory) hearing loss yield contralateral and ipsilateral acoustic reflex thresholds at approximately the same intensity levels as those with normal hearing. Acoustic reflexes are absent in the presence of a severe or profound hearing loss. A significant conductive hearing loss typically eliminates the response on either ear whenever the affected side is stimulated. This is because the stimulating sound is not loud enough to trigger the reflex when the affected ear is stimulated, and the middle ear abnormality (eg, otosclerosis or middle ear effusion) prevents the stapedius muscle from contracting even when the opposite (normal) ear is stimulated. Therefore, any disorder of the stapedius muscle can also cause absent acoustic reflexes. Thus, the only reflex that will occur for a unilateral conductive loss is the ipsilateral reflex to the normal ear. A bilateral conductive loss eliminates the reflex in all four conditions. A lesion of cranial nerve VIII can eliminate both the contralateral and ipsilateral acoustic reflexes whenever the affected side is stimulated (ear effect). However, contralateral and ipsilateral reflexes are usually present when the normal ear is stimulated. For pathologies affecting the central crossed pathways, reflexes are present in both ipsilateral conditions, but may be absent in the two contralateral conditions. A lesion of the seventh nerve (eg, Bell palsy) can eliminate the acoustic reflex whenever the affected side is measured (probe effect), regardless of which ear is stimulated. This pattern can be distinguished from the conductive pattern because in a conductive loss, both the measured and the stimulated ear typically show absent contralateral reflexes. In seventh nerve pathology, the acoustic reflex also can help to determine whether the lesion is proximal or distal to the branching of the stapedius muscle. If the lesion is proximal to the stapedius muscle, acoustic reflexes are absent; if the lesion is distal to the muscle, reflexes are present. Any central nervous system depressant, including alcohol, can depress the amplitude of the response. A functional hearing loss may be suspected if reflexes occur below or at the volunteered pure-tone thresholds. When hearing loss exceeds 70 dB HL, it becomes difficult to determine whether absent reflexes are due to a cochlear or a retrocochlear hearing loss.

Acoustic Reflex Decay

The measurement of acoustic reflex decay may be useful when a retrocochlear lesion is suspected. For this procedure, a 500- or 1000-Hz signal is presented to the ear contralateral to the probe ear, at 10 dB above the patient's acoustic reflex threshold for 10 seconds. If the response amplitude decreases more than 50%, it is considered abnormal (positive) and suggestive of a lesion on cranial nerve VIII. Usually, acoustic reflex decay is only measured contralaterally at 500 or 1000 Hz, because higher frequencies and ipsilateral stimulation may show decay even in normal subjects. The sensitivity and specificity of reflex decay are not as high as the auditory brainstem response (ABR) for identifying vestibular schwannomas. Also, care must be taken in deciding whether or not to administer decay testing, especially in patients with tinnitus or hyperacusis, because of the intense stimulus levels that are often required.

Auditory Brainstem Response

Auditory brainstem response (ABR) testing objectively assesses the neural synchrony of the auditory system from the level of the eighth nerve to the midbrain. The obtained results can be extrapolated to provide information regarding hearing sensitivity and also can be used for neurodiagnostic purposes. To administer the ABR test, electrodes are placed on the patient's head, and a series of sounds are presented. The patient must remain still and, with children, it may be necessary to sedate to obtain valid results. The diagnostic (not screening) test procedure typically requires 30–60 minutes. The ABR consists of a series of 5–7 waves that occur within the first 10–15 ms following the stimulus. These potentials are shown in Figure 45–5. Waves I and II are thought to be generated in the eighth nerve of the stimulus ear; Waves III–V are thought to be generated in the brainstem and midbrain. For neurologic purposes, the latencies (absolute, interwave, and interaural) and amplitudes (absolute and relative) of Waves I, III, and V are analyzed. Clicks are the most commonly used stimuli because their abrupt rise time and broad spectrum enhance neural synchrony; however, the results are dominated by the high-frequency region.

The other primary use of electrophysiologic measures is to estimate the auditory thresholds for AC and BC in patients who are unwilling or unable to provide accurate behavioral audiometric thresholds. For these patients, the lowest intensity level at which Wave V can be visualized and repeated is considered the threshold. ABR is frequently used for newborn hearing screening, because it provides accurate information in a relatively short amount of time. For better definition across the frequency range, frequency-specific stimuli, such as tone pips, are used. Because tonal signals have a slower rise–fall time than clicks, the wave morphology may be degraded, making threshold identification more difficult. Normative values from research centers for gender and age are available; however, to be accurate, each clinic should establish its own equipment-specific normative values. More recent additions to the electrophysiologic battery include ASSR (Auditory Steady State Response), which allows for a potentially more rapid means of establishing frequency-specific thresholds, stacked ABR (a more sensitive and specific method for the detection of small tumors), and CHAMP (cochlear hydrops analysis masking procedure) for detection of Meniere disease.

ABR Interpretation

Interpretation of the ABR depends on knowledge of the relevant recording and subject variables. Age and gender, as well as the type, degree, and configuration of the hearing loss, may substantially affect the latencies and amplitudes of the ABR. The following statements summarize the expected outcomes:

Normal Hearing

The latencies for Waves I, III, and V are within the normal range, and the interaural (between ears) latencies are equal (within 0.2–0.3 ms).

Conductive Hearing Loss

The absolute latencies of all waves are prolonged, but the interwave latencies are not substantially affected. This pattern occurs regardless of the stimulus intensity. The configuration of the hearing loss may also contribute to variability in the latencies.

Cochlear Hearing Loss

The degree and configuration of the hearing loss may affect the latencies of Waves I, III, and V. A relatively flat hearing loss of less than 60 dB HL should not impact the ABR latencies, but high-frequency hearing loss may reduce the amplitudes and prolong the absolute latencies of the waves, without increasing the interwave I–V latency difference when stimuli are presented at high intensities.

Retrocochlear Hearing Loss

A variety of effects on the latencies and morphology of the ABR may occur. These may include the absence of waves, prolonged absolute or relative latencies (2 or 3 standard deviations beyond the mean), or prolonged interaural latencies. Wave I may be within normal limits, but the absolute latency for Wave V, and consequently the I–V interwave latency, is prolonged beyond the normal limits.

Electrocochleography

Electrocochleography (ECOG or ECochG) evaluates the electrical activity generated by the cochlea and the eighth cranial nerve, occurring during the first 2–3 ms subsequent to a stimulus. The active electrode is placed in the ear canal, on the tympanic membrane, or through the tympanic membrane on the promontory; the reference electrode is placed on the vertex or the contralateral earlobe or mastoid. The closer the active electrode is to the cochlea, the larger the response. The principal potentials that are evoked through ECOG are the cochlear microphonic, the summating potential, and the compound action potential. Most commonly, only the summating potential and the compound action potential are of interest. The main application of the ECOG is to help determine if a patient has Meniere disease. The amplitude of the summating potential (reflecting activity of the hair cells) is compared with that of the compound action potential (reflecting whole nerve activity). If the ratio is larger than normal (0.3–0.5), it is considered indicative of Meniere disease. The procedure is considered valid only the patient is symptomatic.

Otoacoustic emissions (OAEs) are objective, noninvasive, and rapid measures (typically less than 2 minutes) used to determine cochlear outer hair cell function. It is believed that OAEs are a by-product of the biomechanical motility of the outer hair cells. Emissions are recorded from a small microphone placed in the ear canal via a soft probe. A good probe fit and low ambient and patient noise levels are essential to record OAEs because the OAE response is very small (usually <20 dB SPL).

OAEs are usually detected in the presence of normal or near-normal hearing. Using current measurement protocols, emissions typically are not detected if there is a conductive or sensorineural hearing loss greater than 25–30 dB HL. Although produced in the cochlea, a pathology in the outer or middle ear may obliterate OAE because of the reduction in stimulation signal intensity and also because the signal must travel distally through the middle and outer ear to be measured by the recording microphone.

The three main types of OAE are spontaneous, transient evoked, and distortion product.

Spontaneous Otoacoustic Emissions

Spontaneous otoacoustic emissions (SOAEs) occur even in the absence of a stimulus. SOAEs are the product of a healthy cochlea; however, they are found in less than half of the normal hearing population and, therefore, cannot be used in hearing screening. They have limited clinical use at this time.

Transient Evoked Otoacoustic Emissions

Transient evoked otoacoustic emissions (TEOAEs) occur when the ear is stimulated by transient or brief signals, such as clicks. Even though the stimulus is broadband, the TEOAE can provide frequency-specific data. As the traveling wave progresses through the cochlea, the basal (high frequency) turn of the cochlea is first to be stimulated by the click and responds earliest, followed by the more middle- and low-frequency (apical) portions, thus allowing the response to be analyzed in both the frequency and time domain.

Distortion Production Otoacoustic Emissions

Distortion production otoacoustic emissions (DPOAEs) are produced when two tones of different, but related, frequencies (F1 and F2) are presented to the cochlea simultaneously. F2 is usually 1.21 times the frequency of F1 and is typically presented 10 dB higher in intensity. In response to these two tones, the normal cochlea generates novel tones, called distortion products, at frequencies related to F1 and F2. This is due to the nonlinearity of the healthy cochlea. The information obtained from DPOAE is frequency specific because tonal stimuli are used to generate the response. Although the results are correlated with the audiometric configuration, the relationship is not precise. Figures 45–6A and 45–6B show examples of TEOAE and DPOAE recorded from a patient with normal hearing.

The clinical applications of OAE are significant. Because OAEs are specific to cochlear function, they can be very useful in differentiating between cochlear and retrocochlear lesions in sensorineural hearing loss. Even though OAEs assess cochlear outer hair cell function, the results are often used in predicting the likelihood of hearing impairment. As such, OAE testing is commonly used in newborn hearing screening because of its speed and noninvasive nature. It is also used in confirming pure-tone test results obtained from young children, in patients for whom a functional hearing loss is suspected, for audiometric configuration confirmation, for ototoxic drug monitoring, and in hearing aid candidacy. More recently, OAEs, in conjunction with ABR, can be used in identifying individuals with auditory neuropathy, also termed auditory dyssynchrony.