3
Characteristics and Limitations of the Hearing in Noise Test
This chapter responds to the first item in the Statement of Task:
[T]o identify and describe the salient test characteristics of the Hearing in Noise Test (HINT), which is currently used to determine the functional hearing ability in adults or children with hearing loss treated with cochlear implantation, and provide recommendations as to how to generalize those characteristics into criteria that can be applied to other validated hearing tests for persons with cochlear implants.
Although the chapter provides a description of the salient characteristics of the HINT, the committee believes that recommending how to generalize those characteristics to other validated tests is likely not the most useful question. As will be discussed, the HINT is no longer the most up-to-date or the most useful test for individuals with cochlear implants. This chapter and subsequent chapters will explain the committee’s thinking and provide recommendations for useful criteria that can be applied to validated hearing tests for persons with cochlear implants.
BACKGROUND
The HINT was originally chosen by the U.S. Social Security Administration (SSA) because it was a widely available test. The SSA Listing of Impairments (the Listings) for hearing loss with cochlear implants in adults states:
- If you have a cochlear implant, we will consider you to be disabled until 1 year after initial implantation.
- After that period, we need word recognition testing performed with any version of the Hearing in Noise Test (HINT) to determine whether your impairment meets 2.11B. This testing must be conducted in quiet in a sound field. Your implant must be functioning properly and adjusted to your normal settings. The sentences should be presented at 60 dB HL (decibel hearing level) and without any visual cues (SSA, 2020).
At the time when that guidance was drafted, speech perception in noise for people with cochlear implants was still very poor, which resulted in the HINT sentences being administered in quiet. However, as cochlear implant technology continued to improve, it no longer made sense to administer the HINT in a quiet sound field, and audiologists today believe a more difficult test is needed to minimize ceiling effects.
Because the U.S. Food and Drug Administration (FDA) approved the use of cochlear implants for hearing loss in adults in 1984, the technology and the surgical techniques involved in its installation have improved considerably and have resulted in ever-improving outcomes for individuals with hearing loss. That has resulted in a continued need for updated materials and assessment criteria to match the improvements in cochlear implant devices.
Historically, the Central Institute for the Deaf (CID) sentence lists were used in cochlear implant research and candidacy. However, as the surgical techniques, technologic processing and components of cochlear implants, and audiologic rehabilitation regiments improved, so did cochlear implant outcomes. With that improvement, the relatively easy CID sentences fell out of favor due to ceiling effects exhibited in tests of cochlear implant recipients. In the 1990s the HINT sentences became the favored assessment for cochlear implantation. The HINT was developed by Nilsson et al. (1994) in response to the need for a broader assessment of speech ability and to address the floor and ceiling effects of the speech assessments that were being used in practice at that time (Nilsson et al., 1994).
DEVELOPMENT OF THE HEARING IN NOISE TEST
As noted in previous chapters, the HINT (Nilsson et al., 1994) is an adaptive speech-in-noise test that is composed of 250 sentences divided into 25 lists. The test is adaptive in that the signal-to-noise ratio (SNR) (see Chapter 2) is to be adjusted based on the performance of the participant. The HINT was designed and validated to fix the noise level and
adjust the speech level adaptively based on the listener’s responses. This means that for each sentence correctly repeated, the following sentence is presented at a lower volume, and for each sentence incorrectly repeated, the following sentence is presented at a higher volume. That adaptive tracking often referred to as a one-down, one-up tracking paradigm with the intention of converging on the SNR required for a score of 50 percent correct. The intended purpose of the adaptive style was to protect against ceiling effects such as those experienced by the relatively easier CID sentence lists at the time. Now, with the ongoing improvements in cochlear implants, the HINT is itself subject to the ceiling effects.
The HINT sentences are based on the Bamford-Kowal-Bench (BKB) sentences. The authors revised the 336 BKB sentences to remove British English idioms for an American English audience while maintaining the sentence length. They had native speakers of American English rate the naturalness of the sentences. The resulting sentences were recorded by a male speaker, and mean-squared amplitudes were calculated for each sentence waveform. All waveforms were rescaled to 67 dB to equate the initial presentation levels. In addition, the average long-term spectrum (i.e., average spectrum of all voiced sounds over the period of time of the presented sentence) were equated and used as a benchmark to generate a spectrally-matched (i.e., speech-shaped) noise masker. That form of masking is based on energy masking rather than on informational masking such as a multitalker babble that contains other random speech. Some studies report speech-shaped masking as being more effective (i.e., the sentences are more difficult to understand). This association may be different in native English speakers than in those who speak English as a second language (Hall et al., 2002; Jin and Liu, 2012). Finally, sentence intelligibility was assessed based on phonemic content and word familiarity. To compensate for intelligibility and equate the difficulty across sentences, the authors increased the mean-squared level of the sentence waveform for sentences with below-average intelligibility (i.e., reduced the presented SNR to increase difficulty).
The modified 250 sentences chosen from the initial 336 BKB sentences were divided into 25 lists of 10 sentences that are matched for phonemic distribution. To assess inter-list reliability in quiet and fixed-noise, authors recruited 18 normal-hearing native English speakers who were presented lists of sentences in randomized order in quiet and noise conditions. Based on repeated measures within subjects, the authors reported a 95% confidence interval for 10-sentence lists estimated at ± 2.41 dB in noise and ± 2.98 dB in quiet. Average thresholds for the HINT were 23.91 dB in quiet and 69.08 dB in fixed noise (72 dB SPL [sound pressure level]). The authors further reported that a threshold measurement with a single list took less than 2 minutes during their study (Nilsson et al., 1994).
ADMINISTRATION OF THE HEARING IN NOISE TEST
For HINT administration in the aided condition, the test is designed to be administered with the participant sitting 1 meter from a loudspeaker that is placed directly in front of the individual (i.e., 0 degrees). The target speech stimuli are presented from the speaker at 0 degrees. The background noise was originally designed to be a steady-state, speech-shaped noise and was presented from the same loudspeaker so that the speech and noise were co-located. However, the HINT is also often presented with spatially separated speech and noise so that the noise originates from an angle of 90 or 270 degrees from the direction of the speech (e.g., Vermiglio, 2008; Vermiglio et al., 2020). Speech understanding in conditions with spatially separated speech and noise is less difficult than when speech and noise are coming from the same place because the listener can benefit from spatial release from masking (SRM). SRM is a phenomenon arising from both head shadow and an across-ear comparison of binaural cues—interaural level differences and interaural time differences—although there are additional cues that may play a role, including spatial attention and better-ear listening (Goupell et al., 2016).
For HINT administration in the unaided condition, the speech and noise stimuli are presented via supra aural headphones or insert earphones (Nilsson et al., 1994). Stimulus presentation can be unilateral or bilateral. In cases of unilateral HINT administration, the non-test ear may require masking so that individual ear performance is quantified. In addition to the standard materials and methods used in the development and validation of the HINT, several studies have administered the adaptive HINT via an eight-loudspeaker circumferential array with speech presented from 0 degrees and uncorrelated noise presented from all eight loudspeakers (e.g., Compton-Conley et al., 2004; Valente et al., 2006). In this scenario, speech and noise are both co-located (from the 0-degree loudspeaker) and spatially separated (for the noise from all of the other loudspeakers), presenting a much more difficult listening environment, given the diffuse nature of the noise. The eight-loudspeaker system is generally used in research laboratories and is not typically incorporated into clinical audiology protocols.
SALIENT CHARACTERISTICS OF THE HEARING IN NOISE TEST
The committee’s Statement of Task directs it to detail the salient characteristics of the HINT. As designed, the HINT has the following characteristics (also see Table 3-1):
TABLE 3-1 Salient Characteristics of the Hearing in Noise Test (HINT)
| Characteristic | Description |
|---|---|
| Sentences | The HINT is composed of 250 sentences that are divided into 25 lists |
| Adaptive assessment | The original design of the assessment uses an adaptive model to adjust the speech level to prevent ceiling effectsa |
| Intelligibility of materials | Phonemic content and word familiarity based on American English are balanced across 25 lists of 10 sentences |
| Accessibility across multiple languages | Translated into at least 11 languagesb |
| Speech-spectrum noise | Noise is spectrally matched to the amplitude and frequency response of the recorded sentences |
| Recorded speech by singular speaker | The HINT materials were recorded by a singular male speaker |
| Co-located speech and noise signals | Assessment designed presentation from a singular sound source (i.e., speech and noise come from the same speaker) |
| Quick assessment tool | Each sentence list from the HINT takes approximately 2 minutes to complete |
| Material access | At this time the HINT is difficult to obtain outside of large academic medical centers |
a The intended use may not be consistent with actual use due to fixed-presentation recommendations from the Minimum Speech Test Battery in 1996.
b The clinician presenting the materials must be fluent in the language of administration.
- Uses sentences
- Adaptive assessment
- Intelligible materials
- Translated into multiple languages
- Speech-spectrum noise
- Recorded male speaker
- Designed for co-located speech from the same azimuth
- Relatively quick assessment
- Difficult to obtain
However, just as improvements in cochlear implant technology and surgical techniques warranted a change from CID to HINT sentences, recent improvements in these areas may warrant SSA re-evaluating its current use of the HINT.
THE HEARING IN NOISE TEST AS A TEST FOR INDIVIDUALS WITH COCHLEAR IMPLANTS
After its inception, the HINT became a widely used speech assessment tool. It has been used to assess communication disabilities among adults with hearing loss (Nilson et al., 1994) and in occupational assessments for hearing-critical jobs (Giguère et al., 2008; Soli et al., 2018a,b). A version has been adapted and validated for children (HINT-C) (Nilsson et al., 1996). It has been translated and validated in multiple languages.1 Of note (and discussed in Chapter 2), the administrator of the HINT must be fluent in the language of the assessment. Additionally, one of its common uses has been as a tool for cochlear implant candidacy and outcomes assessment.
In 1996, stakeholders from otolaryngology, audiology, hearing science, and cochlear implant manufacturers recommended a Minimum Speech Test Battery (MSTB) for cochlear implant candidacy and for assessment of post-operative performance. At that meeting, both the consonant–nucleus–consonant (CNC) word test (Peterson and Lehiste, 1962) and the HINT (Nilsson et al., 1994) were recommended to assess patient performance prior to and following cochlear implantation. In breaking with the intended adaptive presentation format of the HINT, the committee recommended that two 10-sentence HINT lists be presented in quiet (65 A-weighted dB [dB A]) and at fixed-level noise with +10, +5, or 0 dB SNR (Gifford and Revit, 2010; Nilsson et al., 1996).
In 2001 a committee from the American Academy of Otolaryngology—Head and Neck Surgery updated the MSTB and recommended the HINT in its intended administration as an adaptive speech in noise test. However, that recommendation of the adaptive HINT was only intended for post-operative evaluation to avoid ceiling effects, while the HINT sentences in quiet at 70 dB SPL were recommended for preoperative evaluation (Gifford and Revit, 2010; Luxford, 2001). In 2011 the MSTB was revisited by the three manufacturers of cochlear implants in the United States, which resulted in the HINT sentences being dropped in favor of the more difficult, multi-talker Arizona Biomedical (AzBio) sentences (Spahr et al., 2012). The AzBio sentences include 660 sentences recorded from two female and two male speakers.
In the United States manufacturers have historically used the HINT sentences in quiet in the assessment of cochlear implant safety and effectiveness data for FDA. The HINT in quiet is a staple of candidacy indication from FDA-approved insert labeling for devices. Manufacturers used
___________________
1 For example, Arabic (Essawy et al., 2019), Bulgarian (Lolov et al., 2008), Cantonese (Wong and Soli, 2005), French (Vaillancourt et al., 2005), Japanese (Shiroma et al., 2008), Korean (Moon et al., 2008), Malay (Quar et al., 2008), Mandarin (Fu et al., 2011; Wong et al., 2007), Norwegian (Myhrum and Moen, 2008), Portuguese (Bevilacqua et al., 2015), Spanish (Weiss and Dempsey, 2008), and Turkish (Cekic and Sennaroglu, 2008).
slightly different criteria with percent correct criteria, aided conditions, and presentation levels all varying across manufacturers (Gifford and Revit, 2010). Notably, the current Medicare criteria for cochlear implantation do not specify an assessment measure or presentation level, but rather provide general criteria of up to 40 percent correct in best-aided listening condition with tape-recorded open-set sentences (CMS, 2020; Gifford and Revit, 2010).
SSA requires audiometric testing for individuals with a cochlear implant. It considers a person with hearing loss disabled until age 5, or for 1 year after initial cochlear implantation, whichever is later. After that period SSA requires speech recognition testing performed with any version of the HINT to determine whether the person’s impairment meets requirements for disability (SSA Listing 2.11B and 102.11B). The testing must be conducted in quiet in a sound field. The individual’s implant must be functioning properly and adjusted to the person’s normal settings. SSA requirements state that the sentences should be presented at 60 dB HL (Hearing Level) and without any visual cues (SSA Listing 102.00B3b). The current SSA use of the HINT dates back to 2008.2
LIMITATIONS OF THE HEARING IN NOISE TEST
Despite its common inclusion in candidacy and outcomes criteria for cochlear implant recipients, recent work has demonstrated that the HINT is limited by its administration, ceiling effects when presented in quiet or fixed SNRs, ecologic validity, and availability. While using the HINT as intended as an adaptive assessment may avoid ceiling effects, the test is commonly used in quiet or in fixed SNRs, as suggested in the previous MSTB documents (Luxford et al., 2001; Nilsson et al., 1996). This administration deviates from the adaptive presentation and results in ceiling effects that prevent appropriate performance monitoring over time (Gifford and Revit, 2010). In a study of 206 adults with hearing loss (156 cochlear implant recipients, 50 hearing aid users), Gifford et al. (2008) assessed performance on monosyllabic word (CNC) and sentence materials (e.g., the HINT, AzBio, BKB-SIN). The study found that 28 percent of the subjects achieved a maximum 100 percent correct score on the HINT when it was presented in quiet. Furthermore, of all the materials presented in quiet in the study, the HINT exhibited the highest percentage of ceiling effects (Gifford et al., 2008). Others have reported similar findings on ceiling effects. For example, a study of 78 adult cochlear implant users found that the HINT suffered from ceiling effects when presented in quiet at 70 and 60 dB SPL (although not when presented at 50 dB SPL or with background noise). In
___________________
2 73 FR 47103.
a retrospective chart review of adult cochlear implant outcomes, Massa and Ruckenstein (2014) also reported that the HINT Sentences test was more likely to suffer from ceiling effects than AzBio sentences.
As described above, the HINT was developed to keep the noise level fixed and adapt to the presentation level of speech. However, keeping the presentation level of speech constant may be a more relevant approach as talkers will be unable to adjust the level of their voice beyond 2–5 dB SNR in the presence of high-level background noise (e.g., Pearsons, 1977; Smeds et al., 2015). Moreover, the current scoring of the adaptive rule requires that the listener correctly repeat the entire sentence (Nilsson et al., 1994). While this has remained the primary adaptive protocol used for HINT administration, Chan et al. (2008) demonstrated the feasibility of using different adaptive rules tracking different points on the psychometric function allowing for 1–3 errors per sentence. The different methods were intended for use with individuals scoring at or near 50 percent in quiet, allowing the administration of the adaptive HINT across a broader range of listeners (Chan et al., 2008). While that remains a feasible method for adaptive HINT administration, the measure still offers limited ecologic validity, given the use of a single male talker, well-articulated speech, and a steady-state background noise. In addition to those issues, most clinics lack the tools necessary to quickly administer and score an adaptive measure within the confines of a clinical audiometric test suite.
In addition to ceiling effects, the lack of availability of the HINT materials has created problems. In historical context, with the conceptualization of the first MSTB in 1996, cochlear implant surgery was performed only at select major medical centers in the United States. Those centers were able to obtain the necessary test materials and had the appropriate equipment set up (i.e., speaker arrays) to perform speech performance assessments. However, due to its exclusion from the most recent MSTB and the fact that it is no longer available for purchase, the HINT is difficult for clinics across the United States to obtain.
REFERENCES
Bevilacqua, M. C., M. R. Banhara, E. A. Da Costa, A. B. Vignoly, and K. F. Alvarenga. 2008. The Brazilian Portuguese Hearing in Noise Test. International Journal of Audiology 47(6):364–365.
Cekic, S., and G. Sennaroglu. 2008. The Turkish Hearing in Noise Test. International Journal of Audiology 47(6):366–368.
Chan, A. S., M. C. Cheung, S. L. Sze, W. W. Leung, and R. W. Y. Cheung. 2008. Measuring vocabulary by free expression and recognition tasks: Implications for assessing children, adolescents, and young adults. Journal of Clinical and Experimental Neuropsychology 30(8):892–902.
CMS (Centers for Medicare & Medicaid Services). 2020. Cochlear implantation. https://www.cms.gov/Medicare/Coverage/Coverage-with-Evidence-Development/Cochlear-Implantation (accessed December 17, 2020).
Compton-Conley, C. L., A. C. Neuman, M. C. Killion, and H. Levitt. 2004. Performance of directional microphones for hearing aids: Real-world versus simulation. Journal of the American Academy of Audiology 15(6):440–455.
Essawy, W., E. Kolkaila, I. Kabbash, and A. Emara. 2019. Development and standardization of new hearing in noise test in Arabic language. International Journal of Otorhinolaryngology and Head and Neck Surgery 5:1501.
Fu, Q. J., M. Zhu, and X. Wang. 2011. Development and validation of the Mandarin speech perception test. Journal of the Acoustical Society of America 129(6):EL267–EL273.
Gifford, R. H., and L. J. Revit. 2010. Speech perception for adult cochlear implant recipients in a realistic background noise: Effectiveness of preprocessing strategies and external options for improving speech recognition in noise. Journal of the American Academy of Audiology 21(7):441–451.
Gifford, R. H., J. K. Shallop, and A. M. Peterson. 2008. Speech recognition materials and ceiling effects: Considerations for cochlear implant programs. Audiology and Neurotology 13(3):193–205.
Giguère, C., C. Laroche, S. D. Soli, and V. Vaillancourt. 2008. Functionally-based screening criteria for hearing-critical jobs based on the Hearing in Noise Test. International Journal of Audiology 47(6):319–328.
Goupell, M. J., A. Kan, and R. Y. Litovsky. 2016. Spatial attention in bilateral cochlear-implant users. Journal of the Acoustical Society of America 140(3):1652–1662.
Hall, J. W., III, J. H. Grose, E. Buss, and M. B. Dev. 2002. Spondee recognition in a two-talker masker and a speech-shaped noise masker in adults and children. Ear and Hearing 23(2):159–165.
Jin, S. H., and C. Liu. 2012. English sentence recognition in speech-shaped noise and multitalker babble for English-, Chinese-, and Korean-native listeners. Journal of the Acoustical Society of America 132(5):EL391–EL397.
Lolov, S. R., A. M. Raynov, I. B. Boteva, and G. E. Edrev. 2008. The Bulgarian hearing in noise test. International Journal of Audiology 47(6):371–372.
Luxford, W. M. 2001. Minimum Speech Test Battery for postlingually deafened adult cochlear implant patients. Otolaryngology—Head and Neck Surgery 124(2):125–126.
Massa, S. T., and M. J. Ruckenstein. 2014. Comparing the performance plateau in adult cochlear implant patients using HINT and AzBio. Otology and Neurotology 35(4):598–604.
Moon, S. K., S. Hee Kim, H. Ah Mun, H. Kyung Jung, J. H. Lee, Y. H. Choung, and K. Park. 2008. The Korean Hearing in Noise Test. International Journal of Audiology 47(6):375–376.
Myhrum, M., and I. Moen. 2008. The Norwegian Hearing in Noise Test. International Journal of Audiology 47(6):377–378.
Nilsson, M., S. D. Soli, and J. A. Sullivan. 1994. Development of the Hearing in Noise Test for the measurement of speech reception thresholds in quiet and in noise. Journal of the Acoustical Society of America 95(2):1085–1099.
Nilsson, M. J., S. D. Soli, and D. J. Gelnett. 1996. Development of the Hearing in Noise Test for Children (HINT-C). Los Angeles, CA: House Ear Institute.
Pearsons, K. S. 1977. Effect of tone/noise combination on speech intelligibility. Journal of the Acoustical Society of America 61(3):884–886.
Peterson, G. E., and I. Lehiste. 1962. Revised CNC lists for auditory tests. Journal of Speech and Hearing Disorders 27:62–70.
Quar, K. T., Z. S. Mukari, A. A. N. Wahab, A. R. Razak, M. Omar, and N. Maamor. 2008. The Malay Hearing in Noise Test. International Journal of Audiology 47(6):379–380.
Shiroma, M., T. Iwaki, T. Kubo, and S. Soli. 2008. The Japanese hearing in noise test. International Journal of Audiology 47(6):381–382.
Smeds, K., F. Wolters, and M. Rung. 2015. Estimation of signal-to-noise ratios in realistic sound scenarios. Journal of the American Academy of Audiology 26(2):183–196.
Soli, S. D., C. Giguère, C. Laroche, V. Vaillancourt, W. A. Dreschler, K. S. Rhebergen, K. Harkins, M. Ruckstuhl, P. Ramulu, and L. S. Meyers. 2018a. Evidence-based occupational hearing screening I: Modeling the effects of real-world noise environments on the likelihood of effective speech communication. Ear and Hearing 39(3):436–448.
Soli, S. D., A. Amano-Kusumoto, O. Clavier, J. Wilbur, K. Casto, D. Freed, C. Laroche, V. Vaillancourt, C. Giguère, W. A. Dreschler, and K. S. Rhebergen. 2018b. Evidence-based occupational hearing screening II: Validation of a screening methodology using measures of functional hearing ability. International Journal of Audiology 57(5):323–334.
Spahr, A. J., M. F. Dorman, L. M. Litvak, S. Van Wie, R. H. Gifford, P. C. Loizou, L. M. Loiselle, T. Oakes, and S. Cook. 2012. Development and validation of the AzBio sentence lists. Ear and Hearing 33(1):112–117.
SSA (U.S. Social Security Administration). 2020. Disability evaluation under Social Security 2.00: Special senses and speech—Adult. In The blue book. Washington, DC: U.S. Social Security Administration.
Vaillancourt, V., C. Laroche, C. Mayer, C. Basque, M. Nali, A. Eriks-Brophy, S. D. Soli, and C. Giguère. 2005. Adaptation of the HINT (Hearing in Noise Test) for adult Canadian francophone populations. International Journal of Audiology 44(6):358–361.
Valente, M., K. M. Mispagel, J. Tchorz, and D. Fabry. 2006. Effect of type of noise and loudspeaker array on the performance of omnidirectional and directional microphones. Journal of the American Academy of Audiology 17(6):398–412.
Vermiglio, A. J. 2008. The American English Hearing in Noise Test. International Journal of Audiology 47(6):386–387.
Vermiglio, A. J., S. D. Soli, D. J. Freed, and X. Fang. 2020. The effect of stimulus audibility on the relationship between pure-tone average and speech recognition in noise ability. Journal of the American Academy of Audiology 31(3):224–232.
Weiss, D., and J. J. Dempsey. 2008. Performance of bilingual speakers on the English and Spanish versions of the Hearing in Noise Test (HINT). Journal of the American Academy of Audiology 19(1):5–17.
Wong, L. L. N., and S. D. Soli. 2005. Development of the Cantonese Hearing in Noise Test (CHINT). Ear and Hearing 26(3):276–289.
Wong, L. L. N., S. D. Soli, S. Liu, N. Han, and M. W. Huang. 2007. Development of the Mandarin Hearing in Noise Test (MHINT). Ear and Hearing 28(Suppl 2):70S–74S.
