Clinical aspects

Description

The purpose of pure-tone audiometry is to assess auditory acuity for each ear as a function of sound frequency and as a function of whether sound is conducted by air or bone. Such information provides essential information on the underlying causes of hearing impairment and is important in selecting therapeutic interventions, including hearing aids or cochlear implants, when necessary.

An audiometer consists of four parts: (1) the oscillator to change the frequency of the sound waves presented to the patient (which the patient hears as a change in pitch); (2) an audio amplifier; (3) an attenuator to control volume or loudness; and (4) headphones worn by the patient for air-conduction testing, or a bone vibrator placed on the mastoid prominence for bone-conduction testing. The amplitude of a tone is commonly presented in alternating descending and ascending patterns to obtain pure-tone thresholds. Amplitude can be varied in 5-decibel increments. The lowest decibel level at which a sound is heard is called the threshold. More precisely, threshold is considered the intensity at which a patient can just detect the signal 50% of the time. The audiogram gives an easy-to-read profile of the person's threshold of hearing relative to normal hearing at each frequency as measured with the aid of an audiometer. Smartphone audiometric screening applications are being developed, which assess both air- and bone-conducted sound (33; 27; 11; 19; 03; 16; 26); smartphone applications, such as the Mobile Based Affordable Screening Audiometer (MoBASA) smartphone application, are suitably accurate for hearing screening of adults and the elderly with moderate or disabling hearing loss (19; 03; 26).

Sound-field testing is a modification of standard audiometric testing commonly used with infants and young children. Instead of headphones, sound is presented via the air-conduction pathway using speakers on each side of the patient. If the child turns toward the correct speaker just after a signal is presented through one of the speakers, this behavior is reinforced by illuminating or animating a toy. Sound-field thresholds cannot be obtained for each ear separately but, instead, reflect hearing in the better ear. These are marked wavy lines on the audiogram. Quick and consistent sound source localization suggests balanced or nearly balanced hearing between the two ears, whereas asymmetric hearing thresholds result in impaired localization.

A tablet-based automated audiometer has been developed as a method for threshold hearing assessment outside of conventional sound booths (35). Tablet-based automated audiometry has been demonstrated to be useful in children from ages 7 to 12 years (28). Results from the tablet device were generally within 10 dB of those determined by conventional audiometry and did not show proportional bias over the testing range (35; 28).

Automated audiometry devices that provide a reasonable approach for sound booths and exam rooms with low-level background noise are available (04). In addition, smartphone-based audiometry may serve as an accurate and accessible approach to hearing evaluations, especially in settings where conventional pure-tone audiometry is unavailable (08; 21). In a meta-analysis of 25 studies with a total of 4470 patients, the overall sensitivity, specificity, and area under the receiver operating characteristic curve for smartphone-based audiometry were 89%, 93%, and 0.96, respectively; the corresponding values for the smartphone-based speech recognition test were 91%, 88%, and 0.93, respectively (08).

Indications

The typical indication for audiometry is a complaint of impaired hearing (by the patient or family members), especially if supported by clinical observations of hearing impairment either in casual conversation or with bedside hearing assessment (eg, perception of finger rub, watch ticking, etc.). Elevated audiometric thresholds are generally associated with self-perceived hearing difficulty (36).

In 2012, the U.S. Preventive Services Task Force concluded that the available evidence was insufficient to assess the balance of benefits and harms of screening for hearing loss in asymptomatic adults aged 50 years or older (Moyer and U.S. Preventive Services Task Force 2012). This recommendation does not apply to persons seeking evaluation for perceived hearing problems or cognitive or affective symptoms related to hearing loss.

Parents underestimate hearing problems in their children, but they more accurately detect hearing loss if it involves speech-related frequencies that is at least of moderate severity or is bilateral (34). In a study of 827 children evaluated by parental assessment, the standard whispered voice test, and pure-tone audiometry, hearing loss (greater than 25 dB) was found in 6% (31). Parent suspicion of hearing loss in their children showed poor accuracy, with a sensitivity of 21% and a Cohen's kappa coefficient of only 0.025 (fair reliability). The whispered voice test at 6 meters distance found a suspected hearing impairment in 807 (49%) of examined ears; however, its sensitivity was only 57%, with a specificity of 52% and Cohen's kappa coefficient of 0.0254 (poor reliability). Given the prevalence of hearing loss in preschool children and the low accuracy of parental suspicion and the whispered voice test, more efficient audiometric hearing screening is needed for children before school begins.

Audiometry is also regularly used to monitor hearing loss in noisy work environments as part of occupational hearing conversation programs. Furthermore, in occupations that involve hearing-critical tasks, including some military jobs, individuals are required to undergo periodic hearing assessments to ensure that they have not developed a significant hearing loss that could impair their ability to safely and effectively perform their jobs (ie, assessments of "hearing readiness"). The typical approach for this has been limited to pure-tone audiograms, but augmenting this with speech-in-noise testing may substantially improve the assessment of the readiness of service members to perform military tasks in noisy environments (30).

Contraindications

Because pure-tone audiometry is a subjective, behavioral test of auditory acuity, it requires patient cooperation. Young children and patients who are confused, agitated, or severely aphasic with impaired comprehension are not appropriate candidates for pure-tone audiometry. With young children, sound-field audiometry may be feasible. Pregnancy is not a contraindication to pure-tone audiometry.

Results

Disorders of hearing can be conveniently divided into conductive, sensorineural, and mixed hearing loss based on pure-tone audiometry, where conductive hearing loss occurs if there are impediments to conveyance of sound through the external and middle ears to the cochlea (eg, impacted cerumen, middle ear effusion) and sensorineural hearing loss results from dysfunction of the receptors themselves (“sensory”) or the afferent neural pathways (“neural”).

With conductive hearing loss, hearing is impaired for air-conducted sounds, but not bone-conducted sounds, producing an air-bone gap on the audiogram. In contrast, with sensorineural hearing loss, hearing is impaired for both air- and bone-conducted sounds, with different audiogram profiles for different causes of sensorineural hearing loss. Presbycusis is typically associated with a downward-sloping high-frequency loss pattern. Noise-induced hearing loss is typically associated with a notched pattern (generally at 4 kHz). Meniere disease is associated with a low-frequency trough pattern. Mixed hearing loss manifests as a combination of conductive and sensorineural hearing loss.

Adverse effects

There are no significant side effects of pure-tone audiometry.

Scientific basis

The results of a hearing test using an audiometer are summarized in an audiogram. The audiogram is simply a graph of auditory acuity by sound frequency that uses standard symbols to reflect each ear tested, whether the sound was presented by air or bone conduction, whether the sound was presented with or without masking, etc.

The commonly stated range of human hearing is 20 Hz to 20 kHz, but the hearing range essential for understanding speech is generally 250 to 8000 Hz. Therefore, most audiograms use a standard set of frequencies in the range necessary for understanding speech, ie, from 250 to 8000 Hz, with the following frequencies typically included: 250 Hz, 500 Hz, 1000 Hz, 2000 Hz, 4000 Hz, and 8000 Hz. However, in 1990, the American Speech-Language Hearing Association suggested inclusion of frequencies in the range of 125 to 8000 Hz (02). Sometimes intervening frequencies are also measured, eg, 3000 and 6000 Hz. Under special circumstances, high-frequency testing is done in the interval from 8000 to 16,000 Hz- and sometimes to 20,000 Hz. Circumstances in which high-frequency audiometry may be useful include screening for ototoxic and noise-induced hearing loss and for evaluation of presbycusis when audiometry is normal in the conventionally tested range (23; 12; 13; 20; 15; 18).

The decibel (dB) measurement unit originated from methods used to quantify signal losses in telephone circuits. Decibels provide a relative measure of sound intensity: the decibel unit is used to measure the intensity of a sound by comparing it with a given reference level (ie, the threshold of normal hearing at the corresponding frequency) on a logarithmic scale. A change in threshold by a factor of 10 corresponds to a 10 dB change in hearing level. One decibel is one tenth of one “bel,” a unit named in honor of the eminent Scottish-born scientist, inventor, and engineer Alexander Graham Bell (1847–1922).

In acoustics, the bel is occasionally used without the deci- prefix, but for clinical use, the decibel is the generally used unit because 1 decibel is the just noticeable difference in sound intensity for the normal human ear. The human ear has an extremely large dynamic range; indeed, the ratio of the sound intensity that causes permanent damage during short exposure to the sound intensity of the quietest sound that the normal ear can hear is at least 1 trillion (10 to the 12th power), which can be conveniently expressed on a logarithmic scale as 12 bels or 120 dB.

Hearing thresholds for clinical purposes are typically presented in dB HL (decibels hearing level), where the hearing level is based on a reference of normal human hearing thresholds for each frequency. Although normal human hearing thresholds are not uniform when measured against an absolute pressure measurement (ie, sound pressure level), the hearing level approach effectively normalizes hearing thresholds so that 0 dB represents the threshold of normal hearing at each frequency.

Various classifications of the degree of hearing loss exist, but one of the most commonly used is given in Table 1 (09). Often, the normal and slight categories are combined into an expanded “normal” category so that “normal” is considered to include dB HL up to 25 dB; in contrast, others consider “normal” to be up to 20 dB HL (05). An audiogram report accompanying an audiogram will often specify the degree of hearing loss for each frequency range or region, eg, “normal below 2 kHz, precipitously slowing to severe loss above 3 kHz.” Nevertheless, this approach may overlook cases with real and functionally important degrees of hearing loss (15).

As the degree of hearing loss increases, the ability to understand speech is increasingly affected (when hearing impairment is symmetric or if asymmetric as a function of hearing in the better-hearing ear). With mild hearing loss, speech comprehension is difficult in a loud environment. With severe hearing loss, much effort is necessary to conduct group conversations. With profound hearing loss, verbal communication is impossible without a hearing aid.

Table 1. Degrees of Hearing Loss

Masking refers to the presentation of noise to the non-test ear to prevent the non-test ear from hearing the signal presented to the test ear. The thresholds obtained with masking are called masked thresholds. Masking is commonly used for bone-conducted sounds because bone conducts sound so effectively that the interaural attenuation (or transcranial transmission loss) of the bone-conducted signal may approach 0 dB (range usually 0 to 5 dB). As a result, unmasked bone-conduction thresholds reflect the thresholds for the better-hearing ear, which may or may not be the intended test ear. An auditory stimulus presented to the test ear stimulates the cochlear of the non-test ear; this is called cross-hearing. Masking is generally used in bone-conduction testing when there is an air-bone gap, ie, whenever the bone-conduction threshold is at least 10 dB better than the air-conduction threshold at a given frequency.

In contrast, the interaural attenuation for air-conduction sound is about 40 dB for signals presented through standard supra-aural headphones and roughly 55 dB with insert headphones (though varying somewhat as a function of sound frequency). Thus, there is much less risk of inadvertently assessing the threshold for a better-hearing non-test ear with unmasked air-conducted signals. For air-conduction testing, the non-test ear may hear the signal when the bone-conduction threshold of the non-test ear is 40 dB better than the unmasked air-conduction threshold in the test ear. A so-called shadow curve is obtained when the thresholds recorded for one ear reflect the non-test ear’s responses. With unilateral deafness, in the absence of masking, even the totally deaf ear will have apparent responses only about 40 dB worse than the bone-conduction thresholds of the non-test ear.

A marked asymmetry in hearing thresholds for the right and left ears would be necessary before masking would be necessary to assess air-conducted hearing thresholds.

Generally accepted conventions for audiogram symbols have been developed (Table 2), although there is no uniform agreement on the use of all symbols (06; 37). To indicate “no response” at the maximum output of the audiometer, a downward-angled arrow is attached to the lower portion of the appropriate symbol and drawn downward at a 45-degree angle to the right for the left ear symbols and to the left for the right ear symbols (01; 02).

Table 2. Standard Audiogram Symbols

Table 3. Other Common Audiogram Report Notations and Abbreviations

Without knowledge of these principles, physicians perform poorly in interpreting audiograms (17). It is essential for neurologists, who deal with disorders of cranial nerves and often have to assess eighth nerve function, to have an understanding of the basic principles of audiometry. Data-driven audiogram classification systems are being developed to greatly simplify audiogram interpretation for nonspecialists (07; 10).

The gold standard of clinical hearing assessment includes a pure-tone audiogram and a word recognition task (14). Word recognition scores decrease dramatically with age and thresholds in patients with sensorineural hearing loss but relatively little in patients with conductive hearing loss (14). Cochlear nerve degeneration is a prime candidate for word score discrepancies (14).