Speech Audiometry

Specimen Collection

Speech audiometry is conducted in 2 separate rooms, with the audiologist and patient in different areas. One room houses the audiometric equipment, including a speech audiometer, which is typically integrated into a diagnostic audiometer. The speech-testing component of the diagnostic audiometer usually features 2 channels with various input and output options. Input devices may include microphones for live voice testing or recorded materials from different playback devices. Output devices in the testing suite can include earphones, ear inserts, bone-conduction vibrators, and loudspeakers.[3] Testing may be performed using earphones, with materials presented to one or both ears or with a bone-conduction vibrator. Presenting speech materials through loudspeakers in a sound-field environment may be particularly beneficial for pediatric testing.

Calibration

Calibrating speech materials with the audiometer’s volume unit meter is essential for ensuring accurate and reliable test results in speech audiometry. This process guarantees that speech sounds are presented at the correct intensity level.

Calibration involves several steps. First, the audiometer must be checked to ensure proper function and the correct mode is used for speech testing. Next, the calibration tone or speech material is played through the audiometer. The volume unit meter is then adjusted so that the input level peaks at 0 dB, ensuring that the speech material is presented at the correct volume. Then, verification is performed using different speech materials to confirm consistency.

Calibration must be conducted before each testing session, as settings can drift over time due to equipment wear, environmental changes, or variations in input materials. Maintaining accurate calibration ensures that speech sounds are delivered at a standardized level, making test results reliable and comparable across different testing environments.[4][5]

Procedures

Speech audiometry involves several techniques.[6] Speech recognition tests typically evaluate a patient's ability to understand and repeat spoken words presented at different intensity levels.

Speech Awareness Threshold

The speech awareness threshold (SAT), also known as the speech detection threshold (SDT), measures the minimum level at which speech can be detected at least 50% of the time. Unlike other speech recognition tests, the patient is not required to repeat words but simply indicate when speech stimuli are present. This test is particularly useful in pediatric populations, as young children may be too young to understand or repeat words. SAT is also beneficial for patients who speak another language or have a neurological speech disorder.

Spondees, which are 2-syllable words spoken with equal emphasis on each syllable (eg, baseball, hotdog, and airplane), are commonly used in SAT testing. These words are easily recognizable and contain enough information within each syllable to allow for reasonably accurate guessing.

For patients with normal hearing or a relatively flat hearing loss, the SAT is typically 10 to 15 dB lower (better) than the SRT, which requires patients to repeat presented words.

Speech Recognition Threshold

The speech recognition or reception threshold (SRT) measures the lowest level at which speech can be identified at least 50% of the time.[7] Spondees are also used for this test. The SRT is valuable for validating pure-tone thresholds because it strongly correlates with the average of pure-tone thresholds at 500, 1,000, and 2,000 Hz. Ideally, the SRT and 3-frequency average should be within 5 to 12 dB of each other.

Performing SRTs involves several steps. First, the audiometer must be adjusted, and appropriate speech materials must be selected. The patient should then receive clear instructions on their required actions. During testing, words are presented at different intensity levels, and responses are documented. Scoring is performed by calculating the percentage of correctly repeated words.

Suprathreshold Word-Recognition Testing

Suprathreshold word-recognition testing assesses the ability to comprehend and repeat single-syllable words presented at a conversational or other suprathreshold level. One key measure in this testing is the WRS, which evaluates a patient’s ability to recognize and repeat words at a comfortable listening level, typically above the SRT. A higher WRS indicates better speech understanding, while scores below expected levels may suggest auditory processing difficulties or sensorineural hearing loss.[8] This test is also used to assess the effectiveness of hearing aids in improving speech recognition.

Sentence Testing

Sentence testing is another approach used to evaluate a patient's ability to hear and understand everyday speech. This method aims to reflect the contextual characteristics of conversational speech, providing a more realistic assessment of how a patient processes spoken language in daily interactions.

Most Comfortable and Uncomfortable Loudness Levels

The most comfortable loudness level (MCL) refers to the intensity at which speech is most comfortably heard, typically 40 to 50 dB above the SRT. However, this sensation level (SL) is often reduced in patients with sensorineural hearing loss. MCL testing is useful for determining the appropriate hearing aid gain for individuals who may benefit from amplification.

The uncomfortable loudness level (UCL) measures the maximum intensity a patient can tolerate for speech. This assessment helps establish the upper limit for hearing amplification and defines the patient’s dynamic speech range, ensuring that amplified sounds remain within a comfortable listening range.

Indications

Speech audiometry is an essential tool in an audiologist's toolkit to assess how effectively individuals comprehend speech. This form of testing is especially beneficial in various clinical situations.

Diagnosing Degree and Type of Hearing Loss

Speech audiometry helps diagnose both the type and extent of hearing loss. This tool provides a functional assessment of how hearing impairment affects speech understanding, offering insights that pure-tone audiometry alone cannot. Consequently, speech audiometry is particularly important in distinguishing between conductive and sensorineural hearing loss. Additionally, word recognition abilities are evaluated to gain a comprehensive understanding of auditory function.

Assessing Hearing Aid Candidacy

Before fitting a hearing aid, one must understand how well a person can understand speech. Speech audiometry helps determine whether a patient is a suitable candidate for hearing aids and assists in selecting and adjusting the devices for optimal performance. This test also helps determine the appropriate gain and maximum output for amplification devices.

Evaluating Auditory Processing Disorders

For individuals suspected of having auditory processing disorders, speech audiometry can uncover speech recognition difficulties that standard pure-tone tests may not detect. These challenges may include difficulty understanding speech in noisy environments, deficits in temporal processing, and issues with dichotic listening.

Monitoring Progressive Hearing Loss

Patients with conditions that cause progressive hearing loss, such as Meniere disease or medication-induced ototoxicity, require regular monitoring. Speech audiometry tracks changes in speech understanding over time, helping guide adjustments to treatments and interventions.

Cochlear Implant Evaluation and Follow-Up

Speech audiometry evaluates the potential benefits of cochlear implants for individuals being considered for this intervention. After implantation, speech audiometry can help in monitoring speech recognition improvements and adjusting the device’s programming.

Differential Diagnosis of Auditory Pathologies

Speech audiometry helps distinguish between cochlear and retrocochlear pathologies. For instance, patients with relatively good pure-tone thresholds but below-average speech recognition scores may have retrocochlear pathology, such as an acoustic neuroma.

Evaluating Speech Understanding in Noise

Many patients struggle to understand speech in noisy environments. Speech audiometry tests, such as the Hearing in Noise Test (HINT) and Quick Sentence in Noise Test (QuickSIN), specifically assess speech comprehension in noise. These tests provide essential insights into real-world hearing difficulties and help evaluate discomfort or tolerance to speech stimuli.

Legal and Compensation Cases

For hearing loss claims in legal or compensation cases, speech audiometry provides objective evidence of how hearing loss affects speech understanding. This information supports claims and helps determine appropriate compensation outcomes.[9][10][11]

Potential Diagnosis

Speech audiometry is valuable in assessing how different degrees of hearing loss affect speech understanding. The impact varies depending on the severity of the hearing impairment.

• Mild hearing loss (26-40 dB hearing level or HL): Individuals struggle to hear faint sounds and understand speech in noisy environments.
• Moderate hearing loss (41-55 dB HL): Understanding normal conversation becomes challenging, especially with background noise.
• Moderately severe hearing loss (56-70 dB HL): Significant difficulty in understanding speech is experienced, and amplification may be necessary for conversations.
• Severe hearing loss (71-90 dB HL): Speech sounds at typical conversation levels may not be heard without amplification.
• Profound hearing loss (91+ dB HL): Visual cues and assistive listening devices become essential, as speech comprehension is severely impaired.[12]

Speech audiometry may be used to diagnose various clinical disorders, including the following:

• Presbycusis [13]
• Ototoxicity [14]
• Ménière disease [15]
• Vestibular (or acoustic) neuroma [16]
• Noise-induced hearing loss [17]
• Otitis media with effusion [18]
• Otosclerosis [19]
• Congenital or inherited hearing loss

Congenital or inherited hearing loss may present with a "cookie bite" pattern on pure-tone audiometry. Conditions associated with this type of hearing loss include Stickler and Waardenburg syndromes and connexin 26 gene mutations.[20][21]

Specific audiology techniques can provide critical insights into speech perception and hearing impairments. By complementing speech audiometry findings, these strategies expand the test's role in diagnosing and managing hearing disorders.

The "speech banana" is a valuable audiology tool that applies speech audiometry results to illustrate the average pitch, volume, and frequency range of speech sounds on an audiogram. This banana-shaped area on the graph represents speech energy distribution, highlighting the frequencies and intensities of common speech sounds. Vowels typically occur at lower frequencies (250-2,000 Hz) and are louder, while consonants, which are softer but crucial for speech clarity, are found at higher frequencies (2,000-8,000 Hz).

By overlaying the speech banana onto an audiogram generated through pure-tone testing, audiologists can predict the degree of gain or loss for specific speech sounds. This tool covers frequencies from 250 to 8,000 Hz and intensities of approximately 20 to 50 dB HL, making it especially useful for diagnosing hearing loss in pediatric patients and evaluating their language development.[22][23]

The WRS helps distinguish between cochlear and 8th cranial nerve disorders. Cochlear disorders generally allow good speech recognition at higher intensity levels, whereas 8th cranial nerve disorders often result in poor recognition even at high intensities.

Speech-in-noise tests assess how well a person understands speech in different auditory environments. The HINT measures the signal-to-noise ratio (SNR) required for 50% sentence recognition in both quiet and noisy settings.[24][25] The QuickSin evaluates SNR loss by presenting sentences with background noise and determining the SNR at which a patient can correctly repeat half of the sentences.[26][27]

Children's speech testing often requires age-appropriate adaptations. Picture identification tasks involve selecting images instead of repeating words. Conditioned play audiometry engages young children in game-like activities to indicate responses to speech sounds. The early speech perception test assesses speech sound recognition in very young or nonverbal children through behavioral responses.

Normal and Critical Findings

A typical audiogram measures thresholds at frequencies slightly broader than human speech, including 250, 500, 1,000, 2,000, 3,000, 4,000, 6,000, and 8,000 Hz. The primary speech frequencies fall between 500 and 4,000 Hz, although normal human hearing ranges from approximately 20 Hz to 20,000 Hz. The arithmetic mean of the air conduction thresholds at 500, 1,000, and 2,000 Hz is calculated for each patient and referred to as the pure-tone average.[28][29] This metric provides a quick and straightforward summary of hearing ability in the speech frequencies but does not account for high-frequency hearing loss.

Speech audiometry is typically included in a comprehensive audiological evaluation.[30] SAT/SDT and SRT are the 2 main types of speech audiometry commonly performed. Although these tests provide different information, both are used in most cases. SAT is similar to pure-tone testing, except that words replace tones, and the patient indicates when a word is heard. SRT involves presenting 2-syllable words with equal emphasis on both syllables, and the patient must repeat the word accurately.

Beyond SAT and SRT, word recognition or speech discrimination testing may also be conducted. In this test, monosyllabic words are presented at 25 to 40 dB above the SRT, which the patient is asked to repeat. A normal score is 80% or higher. If the score declines as word loudness increases, the phenomenon is known as "rollover," which suggests retrocochlear pathology, such as a vestibular schwannoma.[31]

Audiologists use additional verification tools to gather valuable diagnostic information. Suprathreshold testing for the WRS assesses a patient's ability to comprehend and repeat spoken words at an easily audible volume. The Speech Intelligibility Index (SII) quantifies the proportion of speech cues a listener can perceive, ranging from 0 to 1, where 0 means no speech information is audible, and 1 signifies full audibility.

The Count-the-Dots Audiogram visually represents the SII by plotting a patient’s hearing thresholds against the significance of different speech frequencies. The number of dots above the patient’s threshold line indicates the proportion of audible speech cues. This technique involves several steps. First, the audiometer is calibrated, and appropriate word lists are selected. Next, the task is clearly explained to the patient. Words are then presented at a suprathreshold level, eg, 40 dB SL above SRT. Scoring is performed by calculating the percentage of correctly repeated words.

Choosing the right presentation level is crucial. A level that is too high can cause discomfort or distortion, while one that is too low may not be loud enough for an accurate assessment. Ideally, the level should be set at a comfortable listening volume for the patient, typically about 40 dB SL above the SRT.

Interfering Factors

Speech audiometry evaluates a patient's hearing performance by measuring how accurately they repeat a speech stimulus. This assessment considers morphological constraints, which relate to the part of speech of the target word, and syntactic constraints, which involve the structural complexity of the sentence that provides its linguistic context.[32] Linguistic complexity can hinder speech comprehension even when volume or intensity is sufficient, especially in noisy environments. Other interfering factors include auditory sensitivity, developmental differences in speech and language skills, cognition, memory, and attention. Speech audiometry results tend to be more variable in children than in adults.[33]

Audiologists use specific tools to address factors that interfere with speech audiometry results. One such tool is the performance intensity (PI) function for phonetically balanced (PB) words, which graphically represents the relationship between word recognition scores and presentation levels.[34] In patients with cochlear disorders, the performance intensity function typically improves up to a certain point before plateauing. In contrast, individuals with 8th cranial nerve disorders may exhibit a peak followed by a decline in scores as intensity increases, a phenomenon known as rollover.[35] Rollover ratios help quantify this decline and are calculated using the following formula:

Rollover Ratio = (PBmax - PBmin) / PBmax, where PBmax represents the maximum word recognition score, and PBmin is the score at a higher presentation level.[36]

Another method for analyzing speech recognition performance involves comparing whole-word and phoneme scoring. Whole-word scoring assigns a point for each correctly repeated word, while phoneme scoring evaluates accuracy at the phoneme level, providing a more detailed assessment of speech recognition capabilities. When comparing WRS values, critical differences based on the binomial distribution determine whether variations are statistically significant, helping to distinguish normal variability from meaningful changes in speech recognition.

Complications

Speech audiometry does not cause medical complications. However, test limitations such as linguistic complexity, cognitive factors, and background noise can affect accuracy, potentially leading to misdiagnosis or inappropriate management.

Patient Safety and Education

Patient safety and education are fundamental in audiology, as diagnostic or treatment errors can negatively impact patients and their families. Audiologists, otolaryngologists, and other healthcare professionals must collaborate to refine their knowledge and skills, ultimately improving patient outcomes and satisfaction. Key areas of focus for the interprofessional team include clear communication, patient safety education, technological advancements, electronic health record management, patient privacy protection, health literacy promotion, and evidence-based practice. Each of these factors contributes to enhancing patient safety in clinical settings.[37]

Clinical Significance

Speech audiometry is valuable for assessing speech understanding and identifying specific hearing impairments. Researchers and hearing care professionals have increasingly focused on this tool because it can predict, with 85% accuracy, who will successfully benefit from hearing aids. However, speech audiometry does not measure overall satisfaction with hearing aids.[38][39] Proper calibration, accurate test administration, and careful interpretation of results are essential for diagnosing and managing hearing disorders effectively.

Audiologists can provide precise insights into a patient's hearing abilities and challenges through tools such as the SII, performance intensity functions, and specialized testing for children.[40] Speech audiometry has additional clinical benefits, including measuring loudness discomfort levels, such as MCLs and UCLs, which contribute to successful hearing aid fittings, and the Acceptable Noise Level (ANL) test, which evaluates how much background noise a person can tolerate.[41][42]