Opinion | January 2018 Hearing Review
Let’s not fool ourselves anymore; our standard audiometry tests are not fit for purpose.
In the future it seems likely that PTA will remain a useful tool, even if it ceases to be the “gold-standard” hearing test.
In 1879, nearly 140 years ago, Benjamin Ward Richardson1 presented his findings to the Royal Society of London concerning the medical usefulness of a device invented by David Edward Hughes (Figure 1).
“The instrument may be considered to afford the most satisfactory means for testing the hearing power of all persons who can define a sound…The instrument will be of great use to the physician in determining the value of hearing in those who are deaf, and in determining the relative values of the two organs of hearing…The instrument may be used to differentiate between deafness through the external ear and deafness from closure of the Eustachian tube—throat deafness.”
“In another person who was subject to repeated vertigo…the hearing was so defective that although the external ear on each side was clear and the tympanum natural, no sound could be heard below…
“The instrument promises to be of great service in determining the value of artificial tympanums…fine gold [is] the substance for making the most useful and artificial drum.”
“I hope I have related enough to show that the world of science in general, and the world of medicine in particular, is under a deep debt of gratitude to professor Hughes for his simple and beautiful instrument, which I have christened the audiometer.”1
Over 100 years later, the (pure-tone) audiometer is still considered the “gold standard” of hearing tests, and is used to “determine the relative values of the two organs of hearing,” “differentiate between deafness through the external ear and deafness from closure of the Eustachian tube,” and to assist in the management of “repeated vertigo” and perforated “tympanums.”
Pure tone audiometry (PTA) remains the “gold standard”—but not because it doesn’t have limitations. Like many before and since, Crandell et al3 described the biggest hearing complaint of the elderly as being the difficulty of understanding speech in noisy environments. They concluded that PTA was unable to reliably predict this difficulty and that those with similar audiograms often demonstrate different abilities to understand speech in noise. There was general agreement at that time (and for some years before) that some form of speech audiometry was needed, both for the hearing clinic and for research purposes.
Fast forward to today, 30+ years after the need for speech (in noise) audiometry was identified. In a recent paper, Sharma et al4 reviewed 5 of the most popular speech-in-noise tests. The objective was to compare and contrast these 5 tests so as to enable an audiologist or researcher to be able to make an evidence-based selection. In other words, a single “gold-standard” speech audiometry test has yet to be developed. And, despite the claim made in the article, speech audiometry tests have not become a primary tool in the audiologists’ test battery.
In 2015, the International Collegium of Rehabilitative Audiology (ICRA) produced a set of recommendations for the construction of speech tests.5 The aim was a set of guidelines that would standardize the diagnostic tests for measuring speech recognition (in noise or in quiet) and thereby enable test results to be comparable across languages. Two sets of related guidelines were proposed: one for a digit (triplet) test, and another for a (matrix) sentence test. In their review of the matrix sentence test, Kollmeier et al6 commend it for being able to produce “efficient, reliable, and comparable speech reception thresholds across different languages.” The British English version of the matrix sentence test7 has been available for almost 10 years and yet it has not been added to the test battery of the National Health System (NHS) UK audiologist.
Thus, 140 years after its prototype was demonstrated, PTA remains the predominant (and sometimes only) type of “standard” hearing test used in a typical hearing care practice.
Despite their obvious need, why have speech audiometry tests failed to change this? One simple answer to this question is that they fail to meet the requirements any speech-in-noise test should meet. Gatehouse and Robinson8 suggest speech audiometry has two primary purposes:
1) Measurement of the speech recognition problems experienced, and
2) Measurement of the extent to which management or intervention is able to overcome these problems.
Regarding the latter (#2 above), with a few exceptions, speech audiometry tests do not typically support the evaluation of the benefit of hearing aids. It is not possible to conduct “before hearing aids’’ and “after hearing aids” tests. The matrix sentence test, for example, is a “one-ear” test, with speech in noise presented to the person being tested using the single channel (using one of the two headphones).
Regarding the former (#1 above), there is little science to suggest speech audiometry test results properly reflect the problems a person has recognizing speech in noise. The emulating of PTA by using the single channel and one headphone means—just like PTA—what is tested is not something that occurs in the everyday life of the patient. Even with the delivery of recorded sound to both ears, the use of headphones results in its perception inside of the head. Normal hearing involves locating and processing sound streams9 which have their source at points in the three-dimensional space surrounding the head. Both PTA and speech audiometry testing fail to measure the problems someone has in recognizing speech; this is a direct consequence of tests being developed at a time when we knew less than we do now about how human hearing works.
In a recent (2016) paper, Lewald et al10 reflect upon our current understanding of how our hearing works:
“Speech perception in complex acoustic environments is an astonishing capability of the human auditory system. Subjects are able to detect, localize, and selectively attend to a particular speaker of interest even in auditory scenes composed of multiple competing sources and in the presence of reverberation and background noise, as often occur in everyday hearing situations…The interplay of auditory segregation, grouping, and object formation in speech perception mostly involves mechanisms of sound localization.”10
We have two ears to enable horizontal sound localization and this involves a binaural computation involving both the time and level differences of the sounds arriving at the two eardrums. Spectral analysis of the attenuation of sound by the outer ear is used to enable vertical sound localization (and distinction between sounds from the front and from behind). With vertical localization each ear is responsible for its own hemispace.11 Sound object distance estimation uses multiple cues and calculations including the ratio of the energy of the direct and the reverberant signal, known as the direct-to-reverberant ratio (DRR).12 It is therefore essential for speech audiometry be a free-field test with variable background noise, involving the use of both ears. This would also ensure its suitability for the evaluation of hearing aid benefit.
There are some who argue that PTA, even with its known deficiencies, has obviously stood the test of time and will continue to be useful. This may well be true. However, the evidence for PTA not being “fit for purpose” is ever increasing as our knowledge evolves about how hearing and listening actually work. Just a few examples:
- PTA is unable to differentiate between a healthy cochlea and one which has lost 50% of its inner hair cells (IHCs); neuroplastic changes in the auditory pathway (thought to be a reduction of inhibition) have been found to compensate (to a limited degree) for the failing hair cells.13
- PTA measures tone thresholds with no background noise, but human hearing is constantly shifting its dynamic range up and down when real-world background noise levels rise and fall.14
- PTA is of little use to audiologists who try to help those suspected to have auditory processing disorder (ie, hearing problems in noise and other behaviors consistent with hearing loss despite a normal audiogram). There is no universally accepted test for APD. Current British Society of Audiology APD guidelines15 state that there is a critical need for tests that involve more realistic listening situations.
- There is a strong association between hearing loss and cognitive decline. Both NICE (UK) and NIH (USA) identify peripheral hearing loss as a potentially modifiable dementia risk factor.16 The case for fitting hearing aids at the onset of hearing loss is starting to be made,17 and this means a highly sensitive hearing test is needed.
In the future it seems likely that PTA will remain a useful tool, even if it ceases to be the “gold-standard” hearing test. For example, given that perceived tinnitus pitch generally falls within the area of hearing loss,18 then high-frequency PTA (8-20 kHz) is able to assist with tinnitus diagnosis.
So…Where to, Next?
The arrival of a genuine “gold-standard” hearing test cannot come soon enough. Its absence is a significant hindrance to the advancement of effective rehabilitation for those with age-related hearing loss and for children who are struggling, for whatever reason, to keep up with their peers in the slow development of their hearing.19
Recent developments in home cinema have meant that a room can be equipped with numerous loudspeakers capable of producing a 3D sound experience at a relatively low cost (compared to specialized audiometric suites). For example, the Dolby Atmos system can reportedly reproduce up to 128 simultaneous audio objects, each capable of being placed in the 3D space of a room (Figure 2). A “3D Audiology Room” using similar technology would enable not just free-field speech audiometry but the repeatable and realistic emulation of different room types and different background noises. Additionally, one could envision using tests with different talkers, including adults, children, males, and females, as well as talkers who are recognizable as (among other things) quick, slow, happy, sad, angry, excited, or fearful. The inclusion of a large TV, capable of displaying instructions and visual information, might additionally support more real-world situations and the possibility of greater interactive testing with the patient.
Summary
- It has long been accepted that PTA is unsuitable for the reliable prediction of the ability to understand speech in noise. More recent research has shown PTA to be insensitive to the health of IHCs. This brings into question the use of PTA as the basis for the configuration of hearing aids.
- The reliance on PTA is a factor in the failure of audiology to provide a useful, clinic-based evaluation of hearing aid fittings.
- Many speech audiometry tests have been developed, but their usefulness is limited and they are infrequently used in the hearing healthcare clinic.
- The majority of published speech-in-noise research papers reveal the use of “one-ear” speech-in-noise testing—and this significantly limits the use of their findings.
- Recently available technology designed to create a “home cinema experience” will enable the development of a “3D Audiology Room.” This offers the possibility of realistic speech-in-noise testing in a range of emulated room types, and the option of testing both before and after the fitting of hearing aids.
References
-
Richardson BW. Some researches with Professor Hughes’ new instrument for the measurement of hearing; the audiometer. Proc Royal Soc London. 1879;29:65-70.
-
Stephens S. Hughes’ audiometer. Brit J Audiol. 1979; 13 [Suppl 2]:11-12.
-
Crandell CC, Henoch MA, Dunkerson KA. A review of speech perception and aging: Some implications for aural rehabilitation. J Acad Rehab Audiol. 1991;24:121-132.
-
Sharma S, Tripathy R, Saxena U. Critical appraisal of speech in noise tests: a systematic review and survey. Int J Res Med Sci. 2017;5(1):13-21.
-
Akeroyd MA, Arlinger S, Bentler RA, et al. International Collegium of Rehabilitative Audiology (ICRA) recommendations for the construction of multilingual speech tests. Int J Audiol. 2015;54 [Suppl 2]:17-22.
-
Kollmeier B, Warzybok A, Hochmuth S, et al. The multilingual matrix test: Principles, applications, and comparison across languages. Int J Audiol. 2015;54[Suppl 2]:3-16.
-
Hewitt DR. Evaluation of an English Speech-in-Noise Audiometry Test. Southampton, UK: MSc Thesis, Faculty of Engineering, Science, and Mathematics, University of Southampton;2008.
-
Gatehouse S, Robinson K. Speech tests as measures of auditory processing. In Martin M, ed. Speech Audiometry. London: Whurr;1997:74-88.
-
Beck DL, Flexer C. Listening is where hearing meets brain in children and adults. Hearing Review. 2011;18(2):30-35. Available at: https://hearingreview.com/2011/02/listening-is-where-hearing-meets-brain-in-children-and-adults/
-
Lewald J, Hanenberg C, Getzmann S. Brain correlates of the orientation of auditory spatial attention onto speaker location in a “cocktail-party” situation. Psychophysiol. 2016;53(10):1484-1495
-
Grothe B, Pecka M, McAlpine D. Mechanisms of sound localization in mammals. Physiol Rev. 2010;90(3):983-1012.
-
Ghamdan L, Shoman MAI, Elwahab RA, Ghamry NAE-H. Position estimation of binaural sound source in reverberant environments. Egyptian Informatics Jour. 2017;18(2):87-93.
-
Salvi R, Sun W, Ding D, et al. Inner hair cell loss disrupts hearing and cochlear function leading to sensory deprivation and enhanced central auditory gain. Front Neurosci. January 18, 2017;10:621.
-
Asokan M, Williamson R, Hancock K, Polley D. Homeostatic normalization of sensory gain in auditory corticofugal feedback neurons following auditory deprivation. In press.
-
British Society of Audiology. Position Statement & Practice Guidance: Auditory Processing Disorder (APD)[Draft]. April 2017. Available at: http://www.thebsa.org.uk/wp-content/uploads/2017/04/APD-Position-Statement-Practice-Guidance-APD-2017.pdf
-
Livingston G, Sommerlad A, Orgeta V, et al. Dementia prevention, intervention, and care. Lancet. July 19, 2017;390(10113). Available at: http://www.thelancet.com/commissions/dementia2017
-
Hewitt D. Age-related hearing loss and cognitive decline: You haven’t heard the half of it. Front Aging Neurosci. 2017;9:112.
-
Shekhawat GS, Searchfield GD, Stinear CM. The relationship between tinnitus pitch and hearing sensitivity. Eur Arch Otorhinolaryngol. January, 2014;271(1):41-48.
-
Cameron S, Dillon H. Development of the Listening in Spatialized Noise-Sentences Test (LISN-S). Ear Hear. April, 2007;28(2):196-211.
Citation for this article: Hewitt D. Audiometry and its discontents. Hearing Review. 2018;25(1):20-22.
Correspondence can be addressed to HR or Dale Hewitt at: [email protected]
Just read this article as part of a Google search on loss of speech comprehension. I have hearing loss for over 35 years and like many my biggest complaint with HAs are that they are essentially amplifiers and do little for improved speech recognition especially in noisy environment, despite what manufacturers claim.
I have analyzed my own Audiograms since 1993 and find that 750 -1000 dB tracks reasonable well with word recognition. Maybe an algorithm could be suitable within a narrower respond spectrum to better define when word recognition will be projected to be problematic. As an aside I find also that WR in my right ear is poorer than my left, just as you pointed out in the article.
Beyond understanding my own hearing loss, I have been interested in this work because of its impact on the developments for hearing restoration that I am aware of and support. Unless there are simpler proxies for measurement of word recognition then I don’t know how evaluation of potential treatments will be performed other than try it and see.
The restoration projects seem to be pursuing 3 avenues of research – hair cell regeneration, neural improvements and cortex response. Without know which affect word recognition I don’t see how they will measure the effectiveness of the treatment. Since the overriding need is improved speech recognition, I believe there needs to be better tests done invitro to correlate well with in vivo testing.
It would be useful to understand how frequency of tones influence speech recognition. Perhaps subjects with normal hearing could be tested with head phones that could electronically filter various tones to match audiograms that patients with hearing loss experience and see how this affects speech recognition
Hi Dr Hewitt, thanks for your thought provoking article. I am wondering what your thoughts are about speech tests like the LISN which even though performed using headphones,simulates the spatial cues used in free field so the audiologist is able to measure how well a client is able to use these cues. Also with hearing aid algorithms that are able to provide spatial cues to patients it is possible to do some evaluation aided vs unaided in real life setting to evaluate how well a patient can use these cues. On another note, I personally would like to see some routine emphasis in the clinic teaching clients how we hear in noise, and teaching “hearing exercises”, to help them hone in on spatial and other environmental cues. I have developed a sheet of these exercises but have not evaluated them or anything.