16.3.1 Introduction

Pure-tone audiometry (PTA) is the most common test within audiology, apart from newborn hearing screening tests, and forms the backbone of audiological diagnosis and management. PTA is a behavioural test (i.e. responses require the conscious cooperation of the patient) of the threshold of hearing (the lowest level of an acoustic signal that the patient can hear). It is appropriate for patients of approximately 3–5 years of developmental age and older.

Various protocols and guidelines for PTA have been created by different international organisations, such as the American Speech-Language-Hearing Association (1978), British Society of Audiology (2011a), US National Health and Nutrition Examination Survey (National Centre for Health Statistics 2003), Ireland’s Health and Safety Authority (2007) and World Health Organization (2013). A well-structured PTA protocol, if correctly followed, has high test-retest reliability (Stuart et al. 1991), and intra-subject deviation of less than 10 dB is expected in >94% of subjects (Schmuziger et al. 2004). This article gives an overview of the core concepts of PTA which underlie most modern pure-tone audiometry protocols.

16.3.2 Resources

PTA is performed with an audiometer (which generates the sound stimuli), transducers (which deliver the stimuli to the patient) and a patient response button (which the patient presses to indicate that they have heard the sound stimulus). Audiometers can be stand-alone devices but are these days often attached to conventional computers and controlled by software, allowing results to be digitally recorded and connected with patient medical systems.

Air-conduction transducers include supra-aural headphones, typically Telephonics TDH 39 or Beyerdynamic DT 48 (Fig. 16.2), and (occasionally) insert earphones, such as Etymotics ER-3C (Fig. 16.3). Bone-conduction signals are delivered through a dedicated transducer such as Radioear B71 (Fig. 16.4).

Calibration of each audiometer (together with its dedicated transducers) must be performed at least annually (Stage B calibration, defined in ISO 8253-1 (2010)) in order to ensure the accuracy of the stimuli delivered to the patient. Simple checks (also known as Stage A checks) ought to be performed before each session of use to flag any technical issues or drift in the calibration.

PTA is typically performed within a sound-treated room, which reduces both background noise from outside the room and resonance within it. Soundfield audiometry is not covered in this article (see Sect. 16.4 for more details).

16.3.3 Key Principles of PTA

Bone-conduction testing is typically only performed between 500 Hz and 4 kHz, though some centres test up to 6 kHz. Bone-conduction signals can lead to vibrotactile sensation at lower frequencies (Boothroyd and Cawkwell 1970) and be airborne at higher frequencies (Lightfoot 1979) because of poor distortion performance of bone-conduction transducers at both low and high frequencies (Lightfoot 2000; Lightfoot and Hughes 1993). For this reason, precautions should be taken when using bone conduction above 2 kHz, such as covering the test ear with a supra-aural headphone (BSA 2011a). Both air- and bone-conduction stimulations are limited in their maximum output levels (air conduction is limited to approx. 120–130 dB HL output and bone conduction to approx. 60 dB HL output (both varying at different frequencies and dependent on audiometer manufacturer).

16.3.4 Test Procedure

The patient should be clearly instructed to press the response button every time he hears a sound, even if the sound is very quiet. Other responses, such as the patient’s raising their hand or saying ‘yes’, are reasonable if the patient is physically unable to push the button. Adaptations for patients who do not understand their task (e.g. younger children or people with cognitive difficulties) are covered in Sects. 16.4 and 16.24.

Care should be taken to position the transducers accurately in order to avoid artificially lowering the level of the signal delivered to the ear. Supra-aural headphones should rest on the pinnae with the centre of each headphone-loudspeaker in line with the auditory meatus. The bone-conduction transducer should be placed on the mastoid behind the ‘test’ ear (i.e. the ear which is being tested at the time), not touching any part of the ear and with as little hair between the transducer and the skin as possible.

Test Protocol

Most PTA test protocols are based on the modified Hughson-Westlake procedure (Carhart and Jerger 1959), which establishes the ‘ascending’ psychometric threshold by presenting a stimulus to the test ear at a suprathreshold level and reducing the level in 10 or 20 dB steps until no response is present (i.e. the signal level is too low for the patient to hear). The level is then increased in 5 dB steps until a response is obtained (i.e. until the patient can again hear the signal). From this level the ‘10 down, 5 up’ procedure is repeated until more than 50% of these ascending responses occur at the same level, with a minimum of two ascending trials. As soon as 2/2, 2/3, 3/5, etc., responses have been obtained at the same level, that level can be accepted as the threshold of hearing at that frequency (see, e.g. Fig. 16.5a–d). In practice, it is unwise to pursue individual thresholds for a great number of ascending trials because the patient can become tired and the reliability of their responses can suffer. If no threshold is identified within approx. 5 ascending trials, either start again at a clearly suprathreshold level, move on to another frequency and come back later or take a short break. Different protocols for measuring threshold exist, and results obtained will be slightly different for each one. Consistency of test protocol used is essential in order for results to be truly comparable (Franks 2001).

Presentation of the test tones should be arrhythmic with silences in between; otherwise the patient could (consciously or not) predict the presence of a tone which they cannot actually hear, thus resulting in falsely lowered thresholds. The tones should vary in length, and the patient should be instructed to hold the response button for as long as the sound is present, rather than just pressing it when they first hear the sound. This enables the tester to check the response to both the onset and offset of the tone, thereby increasing the certainty of the response.

This procedure is typically carried out beginning with air-conduction stimuli (starting at 1 kHz) on the suspected better ear. All desired air-conduction frequencies are usually tested on the same ear and then moving to the other ear and then finally using bone conduction.

If air-conduction thresholds are within the normal range (usually accepted as <20–25 dB HL; see Sect. 1.​4 and the ‘Interpretation’ subsection), bone-conduction testing is usually not clinically indicated.

In masking, a noise (usually 1/3-octave narrowband noise, though white noise or others are used in some clinics) is played to the non-test ear at the ‘effective masking level’ (the lowest level of masking noise required to prevent the threshold-level pure-tone signal from being heard). The patient is asked to ignore the noise and only respond when they hear the pure-tone signals (played to the test ear). Using the common ‘plateau’ technique, the levels of noise and signal are increased to the point at which responses occur consistently at a particular signal level despite the noise level increasing further. This is taken as the true threshold.

Masking as part of bone-conduction testing is somewhat more problematic because the inter-aural attenuation of bone-conduction signals can be as low as 0 dB. Bone-conduction masking (in which a bone-conduction signal is masked by noise via an air-conduction transducer) is therefore essential in cases of asymmetrical air-conduction thresholds and where bone-conduction thresholds appear significantly better than air-conduction thresholds (the ‘air-bone gap’). It is, however, clinically inefficient to mask all incidences of an air-bone gap because they are actually more likely than not to occur (Studebaker 1967). Small air-bone gaps should not necessarily be taken as evidence of middle ear pathology (Margolis 2015). The BSA protocol therefore recommends bone-conduction masking only where there is a ≥10 dB gap between the not-masked bone-conduction threshold and the air-conduction threshold of either ear (BSA 2011a).

The tester must be careful not to play uncomfortably loud sound into the non-test ear and must be aware of the risks of central masking and cross-masking. In central masking the patient is unable to identify a tone when masking noise is present. In cross-masking the masking noise is itself cross-heard, with the effect of masking the test ear. In cases such as this, masking to find the true threshold may not be possible.

A more comprehensive guideline for masking is beyond the scope of this article. See specific protocols (e.g. BSA 2011a) for further detail.

16.3.5 Recording the Results of PTA

The Audiogram

The results of PTA are typically displayed on an audiogram (see Fig. 16.6), which is a graph showing auditory thresholds with frequency (Hz) on the X-axis and level (dB HL) on the Y-axis. Different countries and clinics use different symbols, though the circle and cross (representing right- and left-sided air-conduction thresholds, respectively), the use of red and blue (for right and left sides) and results for the right side being displayed on the left side of the page, and vice versa, are universal. Not-masked bone-conduction thresholds are represented by triangles in Fig. 16.6 and masked bone-conduction thresholds by square brackets.

Interpretation of Conductive Hearing Loss

Where bone-conduction thresholds are within the normal range and air-conduction thresholds are outside of the normal range, a conductive hearing loss is indicated (see Fig. 16.7).

Interpretation of Mixed Hearing Loss

Where air- and bone-conduction thresholds are both outside of the normal range but there is a difference between them (an air-bone gap) and bone-conduction thresholds are better than air-conduction thresholds, a mixed hearing loss is indicated (see Fig. 16.8).

Different threshold levels are typically found at different frequencies. For example, otitis media with effusion is typically characterised by conductive hearing loss in the lower frequencies and normal thresholds in the higher frequencies (see Fig. 16.7). Presbyacusis (age-related sensorineural hearing loss) is characterised by normal or near-normal-hearing thresholds in the low frequencies and a sensorineural hearing loss affecting higher frequencies (see Fig. 16.6, right ear).

16.3.6 Limitations of PTA

As with all audiological tests, the cross-check principle (Jerger and Hayes 1976), which encourages the use of other tests (such as tympanometry; see Sect. 16.9) and the subjective impression made by the patient to corroborate the result, is worth following. As the test results rely on behavioural responses, falsely raised thresholds (whether conscious or not) could be obtained. See Sect. 15.​5 for insight into nonorganic hearing loss (NOHL) and Lightfoot and Kennedy (2006) on the use of cortical electric response audiometry (which measures the electrical response to auditory stimulation within the auditory cortex) as ‘objective pure-tone audiometry’. Expectation bias can lead to inaccuracy during testing (Margolis et al. 2015), and there is a risk of responses being falsely interpreted by the tester (especially where time pressure and insufficient training are factors).

Classic pure-tone audiometry only tests the threshold of hearing in response to pure tones. Although this is sufficient for the diagnosis of many audiological conditions, it does not provide full evidence about the patient’s functional hearing (e.g. his/her ability to process suprathreshold speech and other functional sounds or their tolerance of loud sound). See Sect. 16.7 for information on speech audiometry. Loudness investigations are outlined in the following subsection.

16.3.7 Loudness Tests

‘Loudness’ refers to the subjective perception of a sound, whereas ‘sound level’ is an objective measure (see Sects. 1.​3 and 1.​4). Uncomfortable loudness level (ULL) (also known as ‘loudness discomfort level’ (LDL)) testing measures the lowest level at which a sound becomes uncomfortable for the subject. This can be useful in providing additional accuracy regarding the patient’s degree of loudness recruitment (in which a small increase in sound intensity is perceived as a large increase, as is typically found in sensorineural hearing loss) (see Sect. 1.​4), thereby informing hearing aid gain settings, and is also often used in cases of suspected hyperacusis (see Sect. 15.​3).

The basic procedure of ULL testing is to increase the level of a sound stimulus gradually until it just becomes uncomfortably loud for the patient (BSA 2011b). It is important to remember that the ULL is not the same as the threshold of pain. Agreeing an appropriate response method with the patient and issuing clear instructions are obviously crucial. ULLs are usually shown on the audiogram using the and symbols for right and left ears, respectively. Sherlock and Formby (2005) surveyed previous literature and proposed a normal ULL of around of 100 dB HL, but they noted considerable variance in normal ULLs reported between different studies across different signals, frequencies and transducers.

Loudness scaling seeks to investigate the subjective loudness of sound in more detail. Various scales exist, and most use verbal descriptors, e.g. very soft, soft, OK, loud, very loud and too loud (Loudness Growth in 1/2-Octave Bands, LGOB scale) (Allen et al. 1990), and specific test procedures to measure the subjective perception of loudness. The highly subjective and non-linear nature of (loudness) sensation can lead to potentially significant interindividual and inter-scale differences, e.g. Elberling (1999); however the international standard ISO 16832 (2006), based on Brand and Hohmann’s Adaptive Categorical Loudness Scaling (ACALOS) (Brand and Hohmann 2002), has sought to increase reliability. The ‘Würzburger Hörfeld’ and ‘Oldenburger Hörfeld’ loudness scaling tests are well established within German clinics (Lehnhardt and Laszig 2009).

16.3.8 Dead-Region Testing

A cochlear dead region is an area of the cochlea in which the inner hair cells or auditory neurons are damaged to the extent that none are functional (Moore 2001). Given the tonotopic nature of the cochlea and the spread of acoustic sensation over the basilar membrane (see Sect. 1.​4), stimulation of a cochlear dead region by using a signal of a frequency to which the non-functioning hair cells/neurons were tuned results in ‘off-frequency listening’, in which functional hair cells/neurons of a neighbouring region of the cochlea respond. Off-frequency listening is not detected on classic PTA, so specific tests are required. The threshold-equalising noise (TEN) test is one such test (Moore et al. 2000).

In the TEN test, a pure tone is played at the frequency of the suspected dead region, and a specifically designed TEN stimulus is played at 10 dB above the notional threshold of the suspected dead region. If the pure-tone threshold in the presence of TEN stimulus is <10 dB above the level of the TEN stimulus, then no dead region is judged to be present; if the pure-tone threshold is >10 dB above the level of the TEN stimulus, then a dead region is present. If a dead region is identified in a hearing aid user, the gain of the hearing aid could be adjusted to avoid stimulation at the affected frequencies, thus avoiding the distorted perceptual effects of off-frequency listening which can include impaired speech perception (Moore et al. 2000).

16.4.1 Introduction

Behavioural audiometry is an essential part of the diagnosis of hearing loss in very young children, even those younger than 6 months of age. Physiological measures (i.e. tympanometry, otoacoustic emissions, ABR and ASSR) should be complemented with behavioural measures to be able to determine the type and degree of hearing loss. The purpose of behavioural audiometry is to obtain air- and bone-conduction thresholds for the whole frequency range. Cross-checking of all measurements is needed before starting hearing aid fitting. This cross-check principle was described 40 years ago by Jerger and Hayes (1976) and is still up to date.

Behavioural audiometry needs well-experienced audiologists and is a time-consuming measurement that has to be done in optimal conditions. The test protocol will vary with the chronological age of the child. In case of retardation, it follows the developmental age of the child. Therefore you have to know the level of development and physical status (i.e. visual, motor abilities) before starting testing.

Testing can be done in a single or double soundproof booth. In a single soundproof booth, the examiner(s) and the child are in the same room. In a double soundproof booth, two examiners are needed, one with the child and one outside with the audiometer. The choice depends on personal preference. It may be useful to have both options.

16.4.2 Behavioural Observation Audiometry (BOA): 0–5 Months

Although BOA is not generally accepted as an integral part of the paediatric audiometry test protocol, some paediatric audiologists provided evidence of BOA in this target group many years ago (e.g. Delaroche 2001; Delaroche et al. 2004, 2005; Madell and Flexer 2008).

BOA’s purpose is to obtain air- and bone-conduction data as close as possible to the hearing threshold for a broader frequency range of at least 500–4000 Hz.

This can only be performed in optimal conditions and by following a strict protocol.

Between 3 and 5 months of age, unisensorial development becomes intersensory (i.e. the child becomes more able to combine input from more than one sensory system) (Piaget and Inhelder 1969). Most babies are already able to make a head turn to the sound, which can be reinforced by multimodal communication and personalised relationships. This has been described in the Delaroche protocol (Delaroche et al. 2011) as ‘congratulating with smiles, signs, caresses and speech’.

In the following paragraphs, test equipment and setup are explained:

When the child is hungry or does not feel comfortable, it may be too distracted to respond to the quietest sounds it can hear.

Testing has to be done in a soundproof booth with minimal distraction. Minimise tactile or visual impulses; no people other than the parents are allowed. When the mother is feeding, ideally the test can start from the moment the sucking becomes effortless and regular.

Start with bone conduction to obtain the MRL of the better-functioning cochlea without influence of middle ear problems. The vibrator has to be placed only at the mastoid. The diadem of the vibrator has to be adjusted to the size of the child’s head. Continue with insert earphones and be aware that the foam tip is not blocked with earwax (Fig. 16.9). In a later stage, one can use the customised earmould of the hearing aid instead of the foam tip.

Start with bone conduction. Place the vibrator on the mastoid (Fig. 16.9 left), and observe the child’s behaviour for a few minutes before presenting any sound in order to know the behaviour without auditory stimulation. If a bone-conduction hearing loss can be expected, start at the MRL of a normal-hearing child at this age, which will be 30–40 dBHL (Delaroche et al. 2011).

Follow the steps as indicated in the flowchart (Fig. 16.10).

Continue with air conduction using insert earphones (Fig. 16.9 right) on the basis of information from objective measurements, if available, or start from the bone-conduction thresholds you have just measured.

If click ABR testing was performed, start at 2 kHz, 20 dB below the ABR threshold level, and choose the better ear. Preferably it is necessary to obtain ear-specific information as soon as possible. Eventually binaural stimulation by insert phones can be used to have an initial idea about the type of hearing loss and the responsiveness of the better inner ear.

Follow the steps as indicated in the flowchart (Fig. 16.11).

The stimulation and response delay are different for each child. Make sure that this delay remains the same for the same child.

The number of observable responses is limited because of the problem of habituation. With BOA it is not possible to confirm each response as with conditioned behavioural audiometry with older children.

16.4.3 Visual Reinforcement Audiometry (VRA): 5–24 Months

Setup

(Fig. 16.12)

Continue with insert earphones or headphones to determine ear-specific and air-conduction thresholds.

Follow the steps as indicated in the flowchart (Fig. 16.13).

In case of uncooperative behaviour when the vibrator or headphones cannot be used, it is better to work in free field in the first contact. In a second stage, ear-specific information should be obtained.

Free-field measurements are also useful for looking for the functional gain with hearing aids or cochlear implants. Soundfield can also be necessary to give the parents insight into the level of hearing impairment.

In cases where any contact with a vibrator or headphones is refused, it will be necessary to start in a soundfield setting.

Follow the steps as indicated in the flowchart (Fig. 16.14).

16.4.4 Conditioned Play Audiometry (CPA): 2–5 Years

Setup

The infant can sit at a children’s table or in a highchair.

There are several possible test setups for CPA (example in Fig. 16.15). The examiner, preferably, has to sit close (next or in front of) the child in order to control their play quickly and provide social reinforcement.

16.4.5 Hard-to-Test Populations

Testing children with special needs follows the same principles as testing children without additional problems. Some adaptations and specific knowledge may be necessary.

With disabled children, it will be important to keep the carers involved in behavioural tests to be able to choose the right procedure and material and interpret the child’s responses.

16.5.1 Introduction

Tinnitus evaluation may differ according to the time of onset (acute vs chronic), classification (objective vs subjective) and accompanying symptoms; however the following steps are important and performed for almost all cases.

16.5.2 Medical History

An adequate, detailed medical history should gather information about possible causes (such as hearing loss, use of ototoxic drugs including salicylates, quinine and aminoglycosides), otological diseases (sensation of aural fullness, sporadic rotatory vertigo lasting hours or temporarily reduced hearing may give a hint towards Meniére’s disease), non-otological diseases (head trauma, noise exposure and the presence of metabolic diseases symptoms) and the tinnitus percept: date of onset (sudden or gradual), localisation (unilateral, bilateral, in the head), subjective loudness, type of sound (ringing, buzzing, hissing, pulsating), temporal aspects (continuous, intermittent), anything that improves or worsens the tinnitus and impact upon the patient (depression, anxiety, concentration, disrupted sleep). Questionnaires to evaluate the psychological influence of the tinnitus (e.g. the Tinnitus Handicap Inventory (Newman et al. 1996)) are useful.

16.5.3 Clinical Examination

Ear, nose, throat and neck examination is essential for each patient in order to investigate possible objective causes. During this assessment, a temporomandibular joint examination, evaluation of the cranial nerves as well as auscultation of the neck and pressure on the ipsilateral internal jugular vein are very important. If the tinnitus percept is reduced during these manoeuvres, venous aetiology may be indicated (Baguley et al. 2013). In a very small percentage of tinnitus patients, the vascular aetiology may be due to a serious underlying disease such as carotid artery dissection or aneurysm.

16.5.4 Audiological Examination

The basic procedure of pitch matching involves presenting low-level pure tones or narrowband noise of various frequencies and asking the patient to indicate when the pitch is similar to that of their tinnitus. As tinnitus may be a spectrum of sounds, spectral matching is recommendable (Henry 2016).

Loudness matching involves increasing the level of the pitch-matched signal (usually in 1 dB steps) from threshold until the patient indicates that the loudness of the signal and tinnitus are equal. Both procedures should be performed separately for each affected ear if possible.

It is important that the patient understands the difference between pitch and loudness during these matching procedures and that they are informed that slightly inexact responses will not hinder their treatment. The matched frequency and level are often indicated by a ‘T’ sign on the audiogram.

Minimum masking level (MML) procedures are sometimes carried out with the aim of determining the patient’s ability to suppress the tinnitus signal in the presence of competing noise. The level of white noise (a random broadband signal that includes energy at all audible frequencies) is gradually increased until the patient reports any change in the tinnitus (such as it becoming louder, softer or entirely suppressed). Tinnitus matching and MML testing are, however, much less simple than one might expect. For example, many different matching procedures have been tested since the 1940s, and this area is rife with methodological difficulties (see Henry and Meikle 2000).

Tinnitus can be inhibited for up to approx. 1 min in some patients following some types of auditory stimulation. This is known as residual inhibition. According to Henry (2016), studies should be performed to develop optimal stimuli for extending this effect as basis for a new therapeutic approach.

Given that parameters of the tinnitus percept, such as loudness, pitch and quality, do not correlate with the perceived severity of tinnitus (Meikle et al. 1984), tinnitus severity may be better assessed from questionnaires (such as the Tinnitus Handicap Inventory) which address the patient’s perception of their tinnitus and its impact on their lives, rather than psychophysical investigation.

The Tinnitus Handicap Inventory (THI) (Newman et al. 1996) contains 25 self-reporting closed questions related to personal and social situations as well as feelings about tinnitus, designed to assess the severity of the patient’s tinnitus (slight, mild, moderate, severe or catastrophic). Other tinnitus questionnaires exist, such as the Iowa Tinnitus Handicap Questionnaire (Kuk et al. 1990), the Tinnitus Reaction Questionnaire (Wilson et al. 1991) and the Tinnitus Functional Index (TFI) (Henry et al. 2014). But as to our knowledge, there are no standardised questionnaires for children (BSA 2014).

In patients with hearing impairment, the ten-item Tinnitus and Hearing Survey (THS) (Henry et al. 2015) separates complaints caused by hearing loss from self-perceived problems induced by tinnitus by two specific subscales.

Given the strong connection between tinnitus and anxiety/depression, some clinics also use tools such as the Hospital Anxiety and Depression Scale (Zigmond and Snaith 1983) to assess the psychological impact on the patient.

16.5.5 Other Investigations

If, e.g. chronic otitis media or otosclerosis is suspected, high-resolution computed tomography may be useful. Magnetic resonance imaging is usually the preferred method in patients with asymmetric hearing loss or asymmetric tinnitus to rule out a cerebellopontine angle tumour, such as vestibular schwannoma. Magnetic resonance angiography or venography may be necessary in cases with suspected cerebral ischemia, dural sinus thrombosis or neoplastic lesion (Sismanis 2011). Doppler ultrasonography may be useful for exploring the vascular structures in the neck. Blood tests for haemodynamic and metabolic diseases are also important.

16.7.1 Introduction

The assessment of auditory processing and recognition of speech in infants and children is important and complex: the child’s performance in a speech recognition test does not only depend on the auditory skills themselves but also on factors such as language, cognition, articulation and speech, visual cognition, attention and concentration. Typically, these factors are dynamic and changing over time.

16.7.2 Phoneme/Syllable Level

The Ling 6 sound test contains the phonemes for [m], [ah], [oo], [ee], [sh] and [s], being familiar speech sounds that cover the relevant speech range from 250 to 8000 Hz (Ling 1989). It can be used with children who are conditioned for and familiar with this task, by professionals as well as by parents. The aim is to assess or verify the detection or identification of the six speech sounds. With the purpose of assessing detection thresholds, a Ling sounds-based PTA paradigm can be used. The ‘Ling Audiogram’ result (see Fig. 16.16) will show the six thresholds for detection-identification of Ling sounds and allows a quick and easy judgement of whether or not the six phonemes will be audible in normal running speech.

The Ling test uses isolated phonemes. A phoneme-based test, supplying useful information on the basic listening level, can also be based on a syllable form or even by using words.

As an example, in the A§E test (Govaerts et al. 2006), syllables are used to assess detection and discrimination skills in the child. For both skills, an operant conditioning protocol is used: visual reinforcement audiometry (VRA). The response of the child to the detection of a stimulus or the discrimination of a stimulus change has to be reliably observed by using paradigms that are adapted to the developmental level of the child. For the difficult-to-test younger children, VRA (Coninx and Moore 1997) might be the only successful procedure; for older children the usual techniques of play audiometry can be used. With the VRA paradigm, not only phoneme detection but also phoneme discrimination can be investigated. In case of discrimination testing, the visual reinforcer is presented in relation to a sudden change in a sequence of ‘stable’ phonemes or syllables (e.g. da-da-da…) to another sequence of ‘stable’ stimuli (e.g. ta-ta-ta…). A response of the child to the sudden phoneme change is visually reinforced.

For older children, easier and faster ‘same-different’-based tasks can be used instead. The stimuli would be two isolated phonemes (/s/-/f/ or /a/-/e/), two consonant-vowel syllables (/sa/-/fa/ or /ba/-/be/) or a minimal word pair, differing only in the same phoneme contrast (seat-feet or ball-bell). Various responses from the child can be used: verbal feedback (‘same’ versus ‘different’) or pointing to a picture representing same or different. Isolated phonemes or consonant-vowel syllables are considered to be better than words, because the test result for word pairs might depend on the vocabulary of the child. This is caused by lexical neighbourhood processes (Janse and Newman 2013). Examples of word-pair tests are H-LAD (German, Brunner et al. 1998) and WADT (English, Ross 1979).

16.7.3 Word Level

The most frequently used form of speech audiometry in children is based on words. Children develop skills to identify or recognise words earlier than full sentences, as well as earlier than isolated phonemes (a subskill of phonological awareness). For test applications, the words have to be chosen carefully in order to be within the vocabulary of children in the target group age. The words can be presented as stimuli in a quiet or noisy environment. The response of the child can be different: repeating the spoken word or pointing (or comparable action) to a picture or object that represents the spoken word.

A well-known version of such a test is the MTS test (Erber and Alencewicz 1976). Each of the 12 words used as stimulus is in one of three prosodic categories: monosyllables, trochees and spondees. Within each of the categories, the child can identify the words only on phoneme-based information. A picture table is shown (see Fig. 16.17) and the child has to point; as a consequence, the MTP test is a closed-set test. Results can be analysed and differentiated at the prosodic and segmental levels. These closed-set tests are available in many languages.

Open-set tests are similar to many tests used for adult listeners. This choice has to be carefully considered, as interference with the speech and articulation skills and the vocabulary of the child will interfere with the auditory skills as such. A gold standard for an open-set test—for English—is the PB-K test (Haskins 1949), which uses a phonetically balanced word list for children in kindergarten.

Closed tests as described above are used to assess a word recognition score at a selected fixed intensity, at 65 dB SPL or at the individually different most comfortable level (MC).

In order to find a speech recognition threshold (SRT), the test has to be repeated several times; alternatively, adaptive paradigms have to be used. A typical example is the adaptive auditory speech test (AAST) for children aged 3–4 years and above (Coninx 2005). Only six spondee words are used, requiring little vocabulary and minimised influence of working memory and visual cognition when the child is searching for the correct picture to click. When using touchscreen technology, even young children 3–4 years old can do the test and are highly motivated and concentrated. After typically 1–1.5 min, the SRT measurement is completed. The result of a test has to be interpreted by taking the age-dependent normal values into account. An important advantage of AAST is the multi-lingual applicability with nearly language-independent normal values.

For tests in noise, steady-state noise can be used (more likely related to cochlear dysfunctions) (Smits et al. 2013) or fluctuating noises, such as the International Female Fluctuating Masker (IFFM, Holube et al. 2011). The thresholds in IFFM are more likely to be related to retrocochlear dysfunctions, i.e. auditory processing disorders.

Future development in speech audiometry for children will increasingly include modern technology with a touchscreen. This not only increases the level of motivation but also opens options for remote screening and testing.

16.8.1 Diagnosis

There is no universally agreed diagnostic procedure for auditory processing disorder (APD). Diagnostic recommendations have been made by the American Speech-Language-Hearing Association (ASHA 2005a, b) and by the American Academy of Audiology (AAA 2010) on the basis of several distinct forms of impaired auditory processing outlined below. These recommendations are a quasi-international standard, but in practice the recommendations are adhered to only loosely in the USA (Emanuel et al. 2011) and even less so in other countries (Hind et al. 2011). In Germany, the diagnosis of ‘auditory processing and perception disorder’ (APPD) (German: auditive Verarbeitungs- und Wahrnehmungsstörung (AVWS)) includes auditory-specific cognitive impairment as a designated form of APPD (GSPP 2014), but only if that impairment is judged not to be described by other, more widely accepted terms (e.g. attention deficit disorder, language impairment). Some countries have issued practice guidelines in recent years, rather than stipulating comprehensive diagnostic measures and criteria. For example, Australian Hearing has implemented standard procedures for the diagnosis, assessment and management of some aspects of APD (Hearing 2013), and the British Society of Audiology (BSA APD SIG 2011a, 2018) has provided an evidence-based commentary to guide practice.

An attempt to reach a consensus on APD definition, diagnosis and management was the principal aim of two ASHA Technical Reports (ASHA 1996, 2005b). Auditory processing was determined to be the neural mechanisms underlying auditory discrimination and pattern recognition; temporal, spatial and binaural aspects of hearing; and perception of competing and degraded auditory signals. APD was defined as a disorder of one or more of these processes, which were noted to be necessary for both non-verbal and verbal perception. Impairments should be diagnosed by using a variety of non-standard indices (e.g. case history, observation of behaviour, questionnaires) and standardised auditory measures that use both non-verbal and verbal stimuli. Electrophysiological testing (e.g. auditory brainstem response, cortical event-related potentials) was also suggested, but details were not provided. Performance of at least two standard deviations (s.d.) below the (standardised) mean on two auditory processing tests, or 3 s.d. on one test, is ‘generally required for diagnosis’. However, both the ASHA Reports and the AAA (2010) Guidelines suggest that the number and specific type of tests used should depend on individual cases. There is no specification of the number of tests that are to be used, and the criteria, like those proposed by other groups, are arbitrary. Statistically, the more tests performed, the more likely an individual will fail two of them. Wilson and Arnott (2013) showed in a large sample of children (n = 150) that diagnosis rates of APD can range from 7 to 93% depending upon which criteria are applied, even using the same test battery.

16.8.2 Differential Diagnosis

Because APD typically presents as problems involving speech perception, most guidelines recommend verbal (speech-based) testing. For example, the commonly used SCAN test battery (Keith 2009) has four core tests that all use verbal stimuli. Verbal stimuli used in APD diagnosis vary from nonmeaningful phonemes (e.g. ‘da’) or vowel-consonant-vowel (VCV) syllables (e.g. ‘ata’) through commonly used short words (e.g. ‘dog’) to longer words and whole sentences. These stimuli all involve a lesser or greater degree of memory and linguistic processing that strong neurological evidence suggests occurs beyond the central auditory nervous system, primarily in the anterior temporal and frontal brain lobes. An ongoing debate concerns whether APD should be diagnosed when only ‘unimodal’ auditory deficits (i.e. deficits specific to hearing) are found (Cacace and McFarland 2013).

To segregate distinctly auditory deficits, non-verbal testing for APD diagnosis is also recommended. Non-verbal tests attempt to isolate those aspects of auditory processing, as listed above (discrimination, pattern recognition, etc.), that are the building blocks of everyday listening. However, there is scant evidence that either these tests or those involving verbal stimuli are very good predictors of the listening problems that people being assessed for APD typically report (Moore et al. 2010; Watson and Kidd 2009). The most common of those problems are, for children, academic difficulties and attentive listening to speech, especially in challenging environments (AAA 2010; Campbell et al. 2012). In addition, listening to both speech in everyday life, and to verbal or non-verbal tests in the clinic, necessarily accesses cognitive resources that are mostly ‘supra-modal’ (not specific to hearing). These resources include attention, memory, fluid intelligence, motivation and emotion. Professional guidelines urge the clinician to ensure that the person being tested is attentive, but clinicians cannot adequately control or measure cognitive influences on perception with current test procedures.

16.8.3 Occurrence of APD (not ADP!) with Other Learning Problems

Only a small proportion of children referred to audiology clinics are assessed for APD, and only a fraction of them receive a diagnosis of APD (Hind et al. 2011; Moore and Hunter 2013). However, around 20% of all children have speech, language, attention, behaviour, reading or intellectual difficulties. A substantial proportion of them perform poorly on tests of auditory processing (Amitay et al. 2002; Ferguson et al. 2011; Miller and Wagstaff 2011; Sharma et al. 2009) or have behaviour that overlaps extensively with that seen in children with APD (Chermak et al. 1999, 2002). Conversely, most children diagnosed with APD have multiple difficulties. Do those children all have APD? The relation between perception and cognition has been debated for well over 100 years, i.e. whether hearing or vision determines intelligence or vice versa (Spearman 1904). Surprisingly, this is still a current debate, for example, in the extensive literature on language and reading difficulties (Ziegler et al. 2005, 2009). The problem faced by hearing specialists is whether to diagnose APD, recognising the non-specific nature of the diagnosis, or to refer a child for a more broadly based assessment. In either case, identifying individual difficulties is the first step towards appropriate treatment.

16.8.4 How Should APD Be Diagnosed?

It has recently been suggested (Dillon et al. 2012) that specific treatment for APD may be an appropriate form of management for some children. These children may be identified by specific tests, for each of which there is a specific treatment, delivered in a hierarchical fashion. A scheme that builds on this model is shown in Fig. 16.18. The aim here is to provide a specific strategy for clinicians, so that a more rational and uniform approach to management may be delivered and the outcome may better describe a more specific problem that the child has.

16.8.5 Current Situation and Way Forward

Several groups around the world have now issued APD statements, guidelines or white papers, including the American Speech-Language-Hearing Association (ASHA 2005a, b), the American Academy of Audiology (AAA 2010), the British Society of Audiology (BSA APD SIG 2011a, b, 2018; Moore et al. 2013), the Canadian Interorganizational Steering Group for Speech-Language Pathology and Audiology (2012), the German Society of Phoniatrics and Paediatric Audiology (Nickisch et al. 2015), the Australian National Acoustics Laboratory (NAL 2015), the Dutch Position Statement (De Wit et al. 2017) and more recently a group in Europe which has published its perspective on APD (Iliadou et al. 2017). As outlined at the beginning of this section, there is no universally agreed diagnostic procedure for APD, but all of the above documents contribute to international debate and better understanding. High-calibre research, alongside international and interdisciplinary dialogue, is imperative for informing future evidence-based practice. The APD MESHGuide, a new evidence based online resource, offers practical guidance (Campbell et at. 2019).

Remediation is considered in Sect. 18.​14.

16.9.1 Introduction

The ability of the tympanic membrane and ossicular chain to transfer sound waves from the outer ear canal through the middle ear is known by a variety of related terms that have specific technical differences but are often used interchangeably in common practice: admittance, compliance, conductance, immittance and susceptance. The opposite (i.e. the ability of the mechanism to resist the transfer of sound) is typically known as impedance or reflectance.

The stiffer the tympanic membrane, the smaller the amount of sound energy admitted through it (low admittance) and the greater the amount of sound energy reflected by it (high impedance). Sound energy is optimally transferred across the tympanic membrane when the air pressure within the middle ear is equal to that in the outer ear canal, which is typically an atmospheric pressure of 0 daPa (decapascals). Most common disorders of the middle ear, such as middle ear effusion or Eustachian tube dysfunction, affect the air pressure within it, thus stiffening the tympanic membrane.

Tympanometry measures admittance at various external air pressures in order to assess the presence of middle ear disorders. Acoustic reflex testing assesses the basic function of cochlear and retrocochlear structures involved in the acoustic reflex arc by measuring the lowest sound level that triggers contraction of the stapedius muscle (a response that stiffens the ossicular chain and tympanic membrane, resulting in measurably reduced admittance).

While tympanometry and acoustic reflex testing are not direct measures of hearing sensitivity, tympanometry especially plays a crucial role in audiological diagnostics, providing further information in cases of conductive hearing loss and functioning as a cross-check to support the findings of other tests. Other acoustic immittance tests exist (e.g. acoustic reflectometry and tests of Eustachian tube function) but are not common in clinical audiology and are not discussed in this article.

The patient does not need to respond actively during acoustic immittance tests and should ideally be still and quiet.

A tympanometer is used for both tympanometry and acoustic reflex testing. Modern tympanometers are often hand-held, wireless devices linked to a computer, but older, larger stand-alone models are still common. The core mechanism is a probe consisting of three tubes: one attached to an air pump, another to deliver a pure tone (the ‘probe tone’) generated by a miniature loudspeaker in the main body of the device and the third a microphone (Fig. 16.19). The probe is inserted into the outer ear canal with an appropriately sized disposable silicone tip to create an air-tight seal with the wall of the ear canal.

International standards for the calibration of tympanometers (ISO 60645-5 2005) must be adhered to.

16.9.2 Tympanometry

In tympanometry, the air pressure in the outer ear canal is varied over the course of a few seconds, usually from a positive pressure of approximately 200 daPa to a negative pressure of approximately −400 daPa at a rate of 50 daPa/s (e.g. British Society of Audiology (BSA) 2013). The rate and direction of pressure change can be varied in many tympanometers, but variations in these parameters play little role in typical clinical testing. During this pressure change, the probe tone (usually 226 Hz) is played constantly into the outer ear canal, and the sound that is reflected from the tympanic membrane is detected by the microphone.

Results and Interpretation

A real-time trace of the inverse of the amount of reflected sound across the gradually changing air pressure is shown on the tympanometer display (admittance or compliance on the Y-axis and pressure on the X-axis) (Fig. 16.20). This graph is known as a tympanogram. The peak of the trace (i.e. the point at which compliance is highest) reflects the point at which air pressure is equal in the outer and middle ear. This can be compared with established normal values (Table 16.3). The volume of air required to fill the ear canal (known as the equivalent ear canal volume (EECV) or just the ear canal volume) is also recorded and can be compared with norms (Table 16.3). The norms shown in Table 16.3 are from the British Society of Audiology (2013). Other norms exist but the principle of interpretation is the same.

Table 16.3

Tympanogram interpretation when using a 226 Hz probe tone

Result			Interpretation	Description
Peak compliance (mL, mmho/L or cm3)	Peak pressure (daPa)	Ear canal volume (cm3)		Commonly used term	Audiogram type (Jerger 1970)
Adults: 0.3–1.6 Children (6m–6y): ≥0.2 (BSA 2013)	Adults: −50 to +50 Children: −200 to +50 (BSA 2013)	Adults: 0.6–1.5 Children: 0.4–1.0 (BSA 2013)	Middle ear function within the normal range	Normal	A
Higher than normal range	Normal range	Normal range	Hypermobile tympanic membrane or ossicular discontinuity	Hypermobile	AD
Lower than normal range	Normal range	Normal range	Reduced tympanic membrane mobility or ossicular fixation	Reduced compliance	AS
Flat	Flat	Normal range	Middle ear effusion	Flat	B
Normal range	Higher than normal range	Normal range	Possible acute middle ear pathology	Positive	–
Normal range	Lower than normal range	Normal range	Eustachian tube dysfunction	Negative	C
Flat	Flat	Higher than normal range	Grommet in situ and patent or perforated tympanic membrane	Large ECV	–
Flat	Flat	Lower than normal range	Occluding wax or incorrectly placed probe	–	–

A classification system is sometimes used when reporting tympanometric results (e.g. Jerger 1970, Table 16.3), but it is preferable to report the actual values obtained. Some commonly used terms for reporting audiogram shape are also indicated in Table 16.3 (Fig. 16.20a–e showing common results of tympanometry).

Many tympanometers also show gradient, which essentially reports the sharpness or roundedness of the peak, but different manufacturers use different algorithms to calculate this (BSA 2013) so interpretation is not always clear. Gradient does not play an important role in typical clinical interpretation guidelines.

The change in air pressure should not be painful or uncomfortable for the patient in the absence of infection or inflammation. Subjects with pressure-related vertigo, such as that resulting from a perilymph fistula in the vestibular system, may feel some dizziness, but other patients typically do not.

Tympanometry can be performed on compliant patients of any age, with a small but crucial change in probe tone frequency for younger patients (see below).

Tympanogram interpretation is different when using the 1 kHz probe tone. The BSA recommends the following steps (see BSA 2013 for more details): first, drawing a straight line between the compliance levels at the positive and negative pressure endpoints of the trace (−400 and +200 daPa) and then determining if the peak is above this line (‘positive peak’), when middle ear function is essentially normal, or below this line (‘negative peak’) or there is no peak, when ear function should be considered abnormal (Fig. 16.21).

Wideband Tympanometry

Wideband tympanometry (also known as 3D or multi-frequency tympanometry) is a comparatively new development in the clinic. A broadband click is used instead of a single-frequency probe tone, allowing the response of the middle ear at many frequencies (typically 226–8000 Hz) to be measured simultaneously. The results are plotted on a 3D tympanogram (Fig. 16.22). The main measure is no longer compliance or admittance (in mmho/L, mL or cm³) but absorption or absorbance, which is the percentage of the incoming sound energy absorbed by the middle ear (i.e. the percentage which is not reflected back to the probe microphone). Wideband tympanometry has potential benefits, such as measuring multiple different-frequency tympanograms simultaneously, the ability to measure without changing air pressure (therefore safer for postoperative ears) and potentially higher diagnostic power for specific pathologies such as otosclerosis (Shahnaz et al. 2009).

16.9.3 Acoustic Reflex Testing

Significantly suprathreshold acoustic stimulation triggers the acoustic reflex, by which the stapedius muscle contracts. The acoustic reflex arc involves acoustic stimulation of the inner hair cells of one cochlea, followed by afferent innervation of the vestibulocochlear (8th) nerve and the ventral cochlear nucleus, where it is sent to both superior olivary complexes and medial nuclei of the facial (7th) nerve, leading to bilateral efferent innervation of the 7th nerve terminating at the stapedius muscles of both ears (Fig. 16.23). The contraction of these muscles stiffens the ossicular chain and therefore stiffens the tympanic membrane, a movement that is measurable using a tympanometer.

The acoustic reflex threshold (ART) is the lowest signal level (dB SPL) from which the acoustic reflex is triggered. This is usually performed at the pressure level of peak compliance for the individual subject, in order that the response obtained is as large as possible. Various stimuli can be used, typically pure tones (500, 1000, 2000 and 4000 Hz) and broadband noise. Normal values are 90–95 dB SPL for tones and 70–75 dB SPL for broadband noise (Margolis 1993). Pathology affecting any of the structures of the reflex arc can lead to an increase in the threshold or obliterate the response. The acoustic reflex can also be used as an objective measure of recruitment in sensorineural hearing loss (Metz recruitment) (e.g. Thomsen 1955).

Ipsilateral measurements involve stimulating and measuring the response in the same ear; contralateral measurements involve stimulating one ear and measuring the response in the other. Combining the results of ipsi- and contralateral measurements can help identify the site of lesion within the reflex arc (e.g. Emanuel 2009). Contralateral measures are also used to avoid test artefacts that can result from interference of the probe and stimulus tones on ipsilateral measurements (Benton et al. 2004).

Esser et al. (1987) proposed comparing acoustic reflex thresholds elicited by pure tones with those elicited by narrowband noises. They found a 10 dB higher pure-tone acoustic reflex threshold in children with auditory processing disorder (APD). Kunze et al. (2016) could not find such differences in children with or without APD. Another test paradigm in occasional use is the reflex decay. In this test, an acoustic reflex is evoked by a sound signal sustained over a longer duration, usually 10 s. The reflex decay is the amount of time during this acoustic stimulation by which the reflex is sustained at a level above 50%. The normal value is ≥5 s. A shorter duration (i.e. a quicker decay) may indicate retrocochlear or other central pathology, such as myasthenia gravis (e.g. Bischoff et al. 1989). This test is, however, not used very often these days because of the prevalence of imaging techniques such as MRI.

Acoustic reflex testing can be performed on patients of any age; however, given that high signal levels are required to evoke a response even in normal-hearing patients, caution should be used, especially in patients reporting sensitivity to loud sounds, such as suspected hyperacusis, patients with lowered uncomfortable loudness levels or patients with tinnitus. Contraindications are otherwise the same as for tympanometry.

16.10.1 Nature and History of Otoacoustic Emissions

Otoacoustic emissions (OAEs) are very low-level sounds that arise during active contraction of the outer hair cells in the organ of Corti. OAEs can emerge spontaneously or be evoked intentionally. The principle of OAE measurement is the detection of these sounds.

OAEs result from the active mechanism of cochlear amplification (see Sect. 1.​4), a function that plays a significant role in the sensitivity and discrimination of hearing (Heinz et al. 2001). Detection of the OAE is therefore an important objective indicator of inner ear function. It does not, however, provide information about the general status of the auditory system.

The history of OAE measurement dates back to 1978, when David Kemp first demonstrated their existence and utility. Since then, several types of OAE, which can be used in the investigation of inner ear function, have been discovered. The most important clinical applications of OAEs are the transient-evoked OAE (TEOAE) and distortion product OAE (DPOAE). The most common use of TEOAEs these days is in newborn hearing screening programmes (see Sect. 17.​1) and of DPOAEs is in audiological monitoring during therapy with ototoxic drugs (see Sect. 14.​9).

16.10.2 Measurement of OAEs

All forms of OAE are based on the same principle: the detection of sounds produced by the cochlea by using a microphone placed in the outer ear canal. The difference between the various OAE methods lies essentially in the stimuli presented and the processing of results.

OAE measurements are widely used in audiological examination. Various pieces of OAE test equipment can be used, from basic screening devices to more sophisticated clinical equipment, depending on the purpose of the investigation. Screening devices are most common in routine practice and are typically user-friendly, to the extent that they can be operated by trained staff with a secondary school level of education. These devices report results simply as ‘positive’ (OAEs are present) or ‘negative’ (OAEs are absent and further audiological investigation is necessary) (Fig. 16.24).

In all OAE tests, the response must be distinguished from the presence of noise. Various strategies are used to aid this process, such as filtering of the response to focus only on the frequency range of interest. Averaging strategies work on the principle that the response occurs within a particular time window following the stimulus but that noise is essentially random. The results of repeated stimulation and measurement (‘sweeps’) can therefore be averaged to reduce the effects of noise. Artefact rejection is also employed in order to reject sweeps that feature recorded sound higher than a particular sound pressure level, thereby omitting transient background noise.

Because OAEs are produced in the cochlea but measured by a microphone in the outer ear canal, the ability to measure them is hindered by the presence of any obstruction in the middle or outer ear. Such obstructions may include amniotic fluid and debris present in the ears of neonates on the first day of life, possibly until the second day of life.

OAEs do not measure the status of the entire auditory system: they may still be present in retrocochlear pathology (e.g. auditory neuropathy).

The reproducibility of OAEs decreases with age, and they are almost never present in any cochlear pathology. The correspondence of age-related ‘normal’ hearing with OAE presence in older age is, therefore, not reliable.

16.10.3 Spontaneous OAEs

16.10.4 Transient-Evoked OAEs

The presence of amniotic fluid or debris in the ears of neonates may mean that the TEOAE cannot be recorded shortly after delivery.

Newborns at risk of retrocochlear pathology, such as auditory neuropathy, should not be tested solely with TEOAEs. Automated auditory brainstem response (AABR) audiometry should be used in these cases.

Clinical devices enable more detailed analysis of the test data for scientific and research purposes. First studies are performed in the clinical application of high-frequency TEOAEs (Goodman 2011).

16.10.5 Distortion Product OAEs

The examination is usually performed and interpreted for individually tested frequencies, though many OAE devices run through a standard battery of tested frequencies. The frequency-specific characteristic of DPOAEs means that manufacturers have been able to place on the market devices with DPOAE-based hearing threshold estimation (DP audiogram) capability. The extrapolation of DPOAE input/output function can be used to estimate pure-tone thresholds up to 8 kHz (Stavroulaki et al. 2001). The clinical importance of DPOAE is nonetheless a subject of further research.

16.10.6 Other Applications of OAEs

Sustained-frequency otoacoustic emissions are acoustic responses to the presentation of an ongoing tonal stimulus. The wide frequency range (0.5–8 kHz) of these emissions enables the study of inner ear function at specific frequencies, whereas TEOAEs and DPOAEs rely on the combined response or interaction of various areas of the basilar membrane. Because they are identifiable even in cases of severe hearing loss, sustained-frequency otoacoustic emissions may be useful as an objective tool for the quantification of hearing loss. Nevertheless, the method is still an object of study and is not routinely used in practice.

The bibliography includes some articles in which other applications of OAE have been studied, such as synchronised SOAE (Jedrzejczak et al. 2008), stimulus-frequency OAE (Ellison and Keefe 2008) and others. The utility of these methods is under study, and their clinical use in routine practice is currently marginal.

16.11.1 Impact of Evoked Response Audiometry on Objective Threshold Audiometry

The gold standard for hearing threshold assessment is pure-tone audiometry (PTA). This subjective examination requires active cooperation of the investigated subject. On the basis of this principle, a significant number of subjects (infants, psychically and mentally diseased persons, malingerers) cannot pass the examination. Objective threshold audiometry is aimed at covering this gap. Among its methods, evoked response audiometry (ERA) plays a most prominent role in newborn hearing screening and in auditory diagnostics of children and their hearing rehabilitation (Dršata et al. 2015). Optimising these methods has been a subject of intensive research, the essential goal of which is finding a method with the best correlation to the behavioural threshold (Stapells 2011).

Different stages of the auditory pathways can be studied with ERA signals. Early latency responses originate from the brainstem (the auditory brainstem response, ABR). Auditory steady-state responses (ASSR) have generators in the brainstem as well as in more central areas (Picton et al. 2003). Middle latency responses come from the primary auditory cortex and long latency response from the non-primary cortex (cortical ERA, CERA).

The more centrally the source of activity can be found, the more the activity is dependent on the arousal of the patient. Nonetheless, we are in a favourable position to measure responses from the brainstem in sleep as well as in narcosis.

16.11.2 Broadband and Frequency-Specific ABR

There are two kinds of ABR measurements: one with broadband stimuli such as clicks, to test the integrity of the auditory pathway and to determine an overall hearing threshold, and the other with frequency-specific stimuli, to estimate the hearing thresholds at these frequencies.

Jewett (1970) and Jewett et al. (1970) first described evoked responses from the brainstem. Therefore, the potentials from the brainstem were named after him, Jewett waves, enumerated I to V or J1 to J5. The sources of these waves were analysed by dipole analysis of the EEG (Scherg 1991). Scherg (1991) showed that wave I is identical with the compound action potential of the auditory nerve. Wave II is generated at the opening of the internal acoustic meatus, the porus acusticus internus. Scherg (1991) found the source of wave III in the efferent part of the ventral cochlear nucleus, still in the ipsilateral hemisphere. Waves IV and V were thought to originate from the nucleus olivaris superior and lemniscus lateralis, on the ipsi- and contralateral sides, respectively.

The conventional click stimulus (Fig. 16.25) excites the whole cochlea, which results in a clear identification of Jewett’s waves. The intensities used in air-conduction ERA reach from threshold up to 100 dB HL and above. A healthy auditory pathway is characterised by latencies of Jewett’s waves that fall within standard ranges. These latency ranges are age-specific. Jewett’s waves appear delayed and altered or can vanish in cases of conductive, sensorineural or retrocochlear neural hearing loss. An increase of the synchrony of neural activation is reached with the use of chirp stimuli (Fig. 16.25), which compensates for the frequency-dependent time delay of sensory perception in the cochlea (Dau et al. 2000; Elberling et al. 2007). The sound of a broadband chirp starts at low frequency and ascends to high frequency. These stimuli evoke higher amplitudes near threshold and therefore a higher signal-to-noise ratio than clicks.

Although the click and broadband chirp stimuli excite the whole cochlea and generate relatively well-synchronised neural responses, these stimuli are not frequency-specific, and so these methods cannot assess the hearing levels at individual frequencies.

Tone bursts (Fig. 16.25) compared with clicks are referred to as providing more frequency-specific results with good correlation with pure-tone behavioural thresholds (Gorga et al. 2006). In a clinical setting, the tone bursts are usually presented at 500, 1000, 2000 and 4000 Hz. To enhance the frequency specificity, the tonal stimuli are masked with notched-noise (so-called notched-noise ABR, Picton et al. 1979). The effect of masking is to prevent neighbouring regions of the cochlear from contributing to the response (Fig. 16.26, review: Stapells and Oates 1997).

Comparable to the application of broadband chirps for overall threshold ABR measurements, narrowband chirps can also be used for frequency specificity, for which a part of the whole chirp is cut out and covers, for example, only one octave (Wegner and Dau 2002). Especially at low frequency, this narrowband chirp is more efficient than tone bursts (Mühlenberg and Schade 2012). Narrowband chirps have been proven to generate a frequency-specific response, with a reduction in the test time in newborns (Ferm et al. 2013).

The reliability and technical mastering of the described methods have resulted in commercially available ABR systems that offer different kinds of stimulation.

16.11.3 Auditory Steady-State Responses

Auditory steady-state responses (ASSR; steady-state evoked potentials, SSEP; amplitude-modulation following response, AMFR) are auditory evoked potentials from the brainstem and more central structures (Beck et al. 2007). The stimuli are usually continuous pure tones, modulated mostly with modulation frequency around 80 Hz in the amplitude alone or both the amplitude and frequency domains (see Sect. 16.12).

16.11.4 Cortical ERA

Cortical ERA (CERA), also called measurement of late auditory evoked potentials (LAEP), is much less sensitive to muscle activity than ABR, with the drawback that it is more dependent upon vigilance and attention. Different kinds of stimuli come into consideration: clicks, tone bursts (frequency-specific with a length of ≥50 ms) or speech signals (consonant-vowel or vowel-consonant-vowel). The examination requires the patient’s alertness or at the most video sedation.

The reliance of CERA on the patient’s mental vigilance makes the method suitable mostly for calm but awake children and adults (Stapells 2002). The observed response pattern is in a time window between 50 ms and 1000 ms and in a frequency range from near 0 Hz up to 600 Hz.

The CERA has its diagnostic value in the examination of the auditory pathway beyond the brainstem up to the cortical areas. The clinical application of CERA is not widespread and is at its beginnings. It can be useful in questions of maturation (after hearing aid use or cochlear implant), auditory neuropathy, auditory processing disorders, psychogenic hearing disorders, autism, lesions of the central pathway or aggravation and simulation (Walger et al. 2014).

16.11.5 Specific Issues of ERA

The measurement is comparable to that of electroencephalography. Four electrodes are fixed: at the vertex (or forehead), both mastoids and a ground electrode at the forehead, cheek, clavicle or neck (Fig. 16.27). The stimuli are presented via headphone, insert earphones or bone-conduction transducer (Fig. 16.28).

Frequency-specific ERA examination is carried out principally in an identical way as for conventional click-evoked ABR. Obviously, the estimation of hearing thresholds at individual frequencies takes a significantly longer time, and feeble responses to the specific stimuli are more sensitive to the ambient noise and measurement conditions than with a conventional click-evoked ABR.

ASSR measurements generally follow a complete automatic protocol. Thresholds at different frequencies as well as in both ears can be measured simultaneously (Korczak et al. 2012). This possibility leads to a shorter measurement time for a complete threshold estimation. Nevertheless, the major advantage of ABR (controlled by an audiologist) over ASSR remains the traceability of the results and the information about the auditory pathway given by the wave pattern above thresholds.

Level-Latency Curves

With the correct identification of the Jewett waves, especially wave V and in addition wave I, information is gained for differential diagnosis. The latencies of the waves t_V and t_I are a function of stimulus intensity: they increase with decreasing intensity. The comparison of these functions with normal ranges helps to identify different pathologies (Fig. 16.29). The latency difference Δt = t_V − t_I is also taken into account.

In conductive hearing loss, the stimulus reaches the cochlea quieter than normal. Therefore, the curves are parallel but shifted to higher intensities. The difference Δt remains normal.

In sensory hearing loss, the curves at high levels are almost normal and prolonged only at levels near threshold. In the case of low-frequency loss, t_V can even be shortened because of the lacking later low-frequency portion of the signal. The curve shape conforms to a defect of the outer hair cells and normal neuronal processing.

In contrast, in retrocochlear hearing loss latency, t_I is normal but t_V is severely prolonged. The wave pattern can also be irregular, e.g. in auditory neuropathy (Schmidt et al. 2012).

The interpretation of the level-latency graphs provides important hints for diagnostics. But be aware: in infants it is more difficult because of maturation of the auditory pathway, so additional diagnostics should be taken into account.

The results of objective threshold audiometry in the hearing rehabilitation process must be continually monitored, also with the help of behavioural methods, as soon as their achievement is possible.

16.12.1 Characteristics of Auditory Steady-State Responses (ASSR)

Registration of the auditory evoked potentials (AEP) of specific frequencies, from the point of view of response correctness, has been and still is inconvenient (long time of stimulus duration—short time of response latency of auditory potentials). Towards the end of the twentieth century (1998), auditory steady-state responses (ASSR) were developed and introduced to the techniques of electrophysiological auditory investigations (Kuwada et al. 1986, Tlumak et al. 2012; review and tutorial: Korczak et al. 2012, Pruszewicz and Obrębowski 2010). The advantage of this method is the possibility of automatic assessment by statistical processing algorithms. Frequency specificity is obtained by the use of pure tones (sine waves) or ‘chirp’-type auditory stimuli (Stürzebecher et al. 2006) as ‘carrier signal’ with frequencies of 500, 1000, 2000 and 4000 Hz, which are modulated by a signal with frequency around 80 Hz. It is possible to use multiple carriers in parallel by modulation with different frequencies around 80 Hz (Lins and Picton 1995; John et al. 1998). This enables time-saving acquisition setups. The sites where the potentials are generated are the brainstem and the thalamus—thus the test is a combination of two registrations: ABR (auditory brainstem responses) and MLR (middle latency responses). In the auditory system, a steady-state response is evoked at the frequency of the modulation. Names for this method are auditory steady-state response (ASSR) or amplitude-modulation following response (AMFR). Figure 16.30 presents a simple diagram of the ASSR production.

16.12.2 Aim of the ASSR Production

The ASSR test is aimed first at producing a simple objective measurement of hearing whose result correlates directly with the audiogram; it is an instrument complying with the principle of audiological diagnostics, the ‘cross-check principle’ according to Jerger and Hayes (1976), in particular in the diagnostics of young children. The automatic analysis of the ASSR is based on a transformation into frequency space, e.g. by a fast Fourier transform (FFT). In the next step, it is evaluated by statistical testing to determine if there is a response to the stimuli at the modulation frequency or not. The methods used can be the calculation of phase coherence values or the F-test (Dobie and Wilson 1996). Phase coherence is calculated if the measured signal has random phases: then there is no response; if in each repetition the phase is the same and related to the stimulus, there is a valid response. In the F-test the energy at the modulation frequency is compared with that of neighbouring frequencies. Different variations and flavours of these algorithms are possible (Picton et al. 2003). Each device manufacturer uses its own estimation algorithm, and usually they do not disclose their methods.

The result of the measurement is not demonstrated as a typical wave form with assessment of registered wave latency, but evaluated as a graph resembling an audiogram. An example of the ASSR test result is presented in Fig. 16.31.

16.12.3 Indication and Interpretation of ASSR

The indication for the ASSR test is the need to obtain an objective audiogram, besides its use in newborn hearing screening. As the ASSR method makes use of tonal, frequency-specific stimuli (similar to that in the audiometric test), the test result can be compared with the behaviourally obtained audiogram. It can be a perfect instrument for objective hearing tests in children and uncooperative patients through earphones and in a free acoustic field through loudspeakers. The problem of this method is the need for a correction to estimate the hearing threshold. In normal-hearing adults and older children, this correction is averaged at 11 dB for 2 kHz up to 17 dB for 500 Hz (Tlumak et al. 2007). In hearing-impaired adults, it varies from 8 to 14 dB. Despite this correction the method is considered to give a good estimate of hearing threshold. In children <6 years and infants, compared with ABR thresholds—which may lie 10 dB above behavioural thresholds—the ASSR for normal hearing is 17 dB, 13 dB, 13 dB and 9 dB worse than ABR in normal hearing and 9 dB, 9 dB, 9 dB and 6 dB worse in patients with hearing loss at 500 Hz, 1000 Hz, 2000 Hz and 4000 Hz, respectively (Van Maanen and Stapells 2010). These corrections vary in different publications (e.g. Chou et al. 2012; Luts et al. 2006; Rance et al. 2005) because of different study groups, stimuli, measurement time, noise floor and other parameters. Considerable variability in interpersonal and intrapersonal results is noticed. A large number of audiologists seem to indicate that so far the method represents poor diagnostic value (Pruszewicz and Obrębowski 2010; Tlumak et al. 2012). Owing to the good correlation of the ABR and ASSR thresholds, the method is quite usable. In commercial ASSR systems, an automatic correction is normally included; therefore the result of measurement is an estimated audiogram. The audiological assistant or semi-skilled examiner does not need to know much about the interpretation of the response, as in click or notched-noise ABR, as the measurement system delivers an audiogram. But this audiogram should be interpreted with caution depending on the degree of hearing loss, age and measurement conditions.

16.15.1 Developmental Balance Assessment in Children: Development of Balance, History and Observation

The vestibular labyrinth and receptors are developed to an adultlike form by 32-week gestation. The vestibular nerve is the first cranial nerve to complete myelinisation. The slow (reflexive) component of VOR can be seen at birth, but the fast (saccadic or central) component is immature and continues to develop until the age of 2 years. Absence of the VOR by the age of 10 months is an abnormal finding (Eviatar and Eviatar 1979; Fife et al. 2000). Infants prefer visual inputs to maintain balance, whereas adults rely more on somatosensory information. Infants perform better under conditions of sensory loss [e.g. visual difficulties] than conditions of sensory conflict, indicating that the central vestibular integrative pathways are immature. The role of the vestibular system in postural control depends on these integrative pathways, and the ability to resolve intersensory conflict is mature approximately at the age of 15 years (Steindl et al. 2006; Shinjo et al. 2007).

Balance difficulties in children may present as delayed motor milestones, poor tone and floppy posture, frequent falls and unusual gait. They may be related to visual, neurological or vestibular pathology or psychological factors. Cardiac pathology and metabolic disturbances such as hypoglycaemia should also be considered. A comprehensive balance evaluation should include a clinical assessment of all the above conditions. The vestibular aspects of a child’s balance can be assessed from history, examination and selected investigations.

16.15.2 Tests for Posture/Vestibulospinal Reflex

The postural tests are limited by lack of specificity for vestibular dysfunction and paucity of normative data for children.

16.15.3 Tests for Eye Movements and VOR

The VOR in children is dependent on the attention and state of arousal of the child, unintended ocular fixation due to light leaks and difficulty with head stabilisation during testing. Examination of children is helped by using child-friendly visual targets (e.g. stickers, toys). Begin by looking for a squint and spontaneous nystagmus. The best way to look for vestibular nystagmus is by using Frenzel’s or Video-Frenzel’s glasses in the central, right and left gaze (angle of gaze should be <30°). Smooth pursuit is best checked with a toy and optokinetic nystagmus with a striped drum with child-friendly pictures.

16.15.4 Tests for Neonatal Reflexes

In the article of O’Reilly et al. (2011), various approaches are presented.

16.15.5 Other Tests: Dix-Hallpike Positional Test

This involves rapidly taking the child from a seated to a lying back position with the head turned to one side (right or left). The eyes are observed for nystagmus. The child is then rapidly taken to the seated position and the eye observation is repeated. The presence of horizontal, rotatory, geotropic nystagmus with vertigo after a short latent period and fatigue on repeated testing occurs in benign paroxysmal positional vertigo (BPPV).

16.15.6 Vestibular Investigations

Adequate preparation for the tests is important. Vestibular sedatives are withheld for 24–48 h prior to testing, and it is important that the child eats only a light meal before the appointment.

16.16.1 Introduction: Relation of Hearing, Language and Speech

According to Aslin and Smith (1988) and Carney (1996), basic sensory perception (sound detection) is followed by coding perceptual representations (phonetic discrimination) and cognitive/linguistic processing (word recognition). These levels are mirrored in speech production by primitive vocalisations, complex vocal utterances (babbling) and phonemic or syllabic speech patterns (Eisenberg et al. 2007). Hearing loss (HL) in early childhood impairs the development of speech recognition and speech production by hindering the generation of correct recognition categories, linguistic and motoric patterns. Recognition of word boundaries, of syllables, of prosodic elements of speech, of sound patterns in words is basically essential for language development in early infancy.

Results in speech recognition and production depend on the type and degree of hearing loss, onset of manifestation, age at detection and start of intervention, additional handicaps, kind and quality of hearing device supply, communication mode, support by parents and intensity of therapeutic training.

Typical findings in hearing-impaired children, dependent on the degree of hearing loss, may be:

• 
  

  Prolonged phase of primitive vocalisations.

  
  
• 
  

  Delay in onset of babbling and differences in babbling behaviour, less consonantal types.

  
  
• 
  

  Delay in development of phonemic/syllabic speech patterns.

  
  
• 
  

  Lagging behind in vowel and consonant production.

  
  
• 
  

  More problems in recognising consonant than vowel contrasts.

  
  
• 
  

  More problems in recognising consonant manner and rear place than consonant voicing and front place.

  
  
• 
  

  Substitution particularly for fricatives and affricates.

  
  
• 
  

  Mild to moderate hoarseness.

  
  
• 
  

  Nasal resonance problems (Eisenberg et al. 2007).

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Kinds of Developmental Disorders of Speech and Language
2. of Developmental Disorders of Speech and Language
3. and Differential Diagnosis of Developmental Disorders of Speech and Language
4. and Prognosis of Voice Disorders

Stay updated, free articles. Join our Telegram channel

Join