Patterns of Aided Loudness Growth in Experienced Adult Listeners with Early-Onset Severe–Profound Hearing Loss

Key Words

hearing aids - hearing loss - prelingual deafness - sensorineural loudness perception

INTRODUCTION

Listeners with bilateral severe–profound hearing loss (S/PHL) pose unique hearing aid (HA) fitting challenges, including restricted dynamic range, worsening frequency selectivity at higher stimulus levels, poorer temporal resolution at lower stimulus levels, and—for many congenitally deaf adults—negligible capacity to recognize speech through listening alone, typically associated with psychoacoustic impairments ([Faulkner et al, 1992]; [Kuk and Ludvigsen, 2000]; [Launer and Kühnel, 2001]; [Kishon-Rabin et al, 2009]; [Kuk et al, 2012]). Cochlear implants are now widely available for this population as an alternative means of accessing acoustic input, with improved outcomes, given advances in speech processing and design of implant components; however, perspectives within the Deaf community are not uniformly positive toward implants, and the ethics and efficacy of implants remain controversial ([Gale, 2011]). Thus, audiologists continue to fit HAs for this population. Advances in digital HA technology and use of evidence-based best practices for HA fitting now enable close approximations to prescription targets in cases of S/PHL ([Keidser et al, 2008]; [Quar et al, 2013]; [Ching et al, 2015]), although a given listener’s preferences along with the severity of the hearing loss can increase the difficulty of the fit for these listeners ([Keidser et al, 2007]).

To suit individual preferences and unique listener characteristics such as those of HA users with S/PHL, clinicians make modifications to prescription settings as an essential and accepted component of best practices ([American Speech-Language-Hearing Association, 2006]; [Valente et al, 2006]; [Keidser et al, 2007]; [Sanders et al, 2015]). There is a risk, however, of compromising prescription fitting goals. In particular, aided loudness perception might be altered because loudness is an important factor in calculating fitting targets in the two commonly used procedures. DSL v5 and m[i/o] (multistage input/output algorithm for digital signal processing [with wide dynamic range compression amplification]) formulas include normalizing loudness across a wide frequency range. National Acoustic Laboratories (NAL)-NL1 and NL2 calculations seek to manage overall loudness and thereby equalize important midrange speech frequencies to enhance speech intelligibility ([Palmer and Lindley, 2002]; [Scollie et al, 2005]; [Johnson, 2012], [2013]). Thus, loudness outcomes can differ with choice of prescription ([Ching et al, 2013]; [Ching et al, 2015]). Without the fine-tuning of a prescription fit, however, loudness is referenced to the average listener with hearing loss under standard listening conditions and thus can bring about unacceptable aided loudness perceptions for a given listener ([Smeds, 2004]; [Scollie et al, 2010]). Work is ongoing to improve the listener’s loudness experience by accounting for sources external to the listener ([Moore et al, 2016]), but internal sources linked to loudness discomfort level and optimal gain setting vary across individuals, even with the same level of hearing loss ([Bentler and Cooley, 2001]). Specific to S/PHL, a prescription fit also may not be able to account for degree of inner hair cell or nerve function, frequency or temporal resolution, or cochlear dead regions, which are all factors that affect loudness perceptions ([Humes, 1982]; [Faulkner et al, 1990]; [Brokx et al, 1997]; [Moore, 2001]; [Vickers et al, 2001]; [Baer et al, 2002]; [Moore, 2004a]; [Kishon-Rabin et al, 2009]; [Malicka et al, 2010]; [Cox et al, 2011]). As a result, predictions of rate of aided loudness growth for a given listener will not always be accurate, regardless of fitting formula ([Johnson, 2013]; [Johnson and Cox, 2013]).

“Restoring” normal relationships between real-world sound intensities and a listener’s perception of the loudness of those sounds also affects the potential for speech communication benefit, part of a long-running conversation about whether normal is best ([Byrne, 1996]; [Moore, 2003]), especially with S/PHL when there has never been an experience of “normal.” Loudness normalization can have a positive outcome for individuals with sufficient auditory function to make use of phonemic speech cues ([Johnson, 2013]); however, there can be a negative effect on speech perception ([Moore, 2007]), especially for listeners with S/PHL, unless there is a careful consideration of HA settings ([Kuk and Ludvigsen, 2000]). There is a wide range of auditory perceptual ability with S/PHL, from detection (merely being aware of the presence of sound), to some level of pattern discrimination skill (e.g., perceiving suprasegmental speech features such as syllabic stress pattern based on the time-intensity waveform), to a progression of phonemic feature recognition (e.g., extracting segmental speech features such as nasality, voicing, or place of consonant articulation) ([Erber, 1979]; [Boothroyd, 1984]; [Geers and Moog, 1992]). This diversity further supports a customized fitting approach with S/PHL HA users ([Ching et al, 2008]; [Schum, 2009]).

Loudness normalization also has implications for HA satisfaction. Both first-time and experienced HA users (the typical study group with mild or moderate-to-severe hearing loss) prefer adjustments that will reduce overall loudness ([Smeds, 2004]; [Smeds et al, 2006]) and modify normalized loudness features that make an HA unacceptably “noisy” ([Moore, 2007]). That is, for them, HA satisfaction is associated with an experience of nonnormalized aided loudness perception ([Shi et al, 2007]; [Johnson and Cox, 2013]). Responses from adults with S/PHL indicate that they are about as satisfied with their overall aided listening experience as listeners with lesser degrees of hearing loss; however, they are much less satisfied with the comfort of loud sounds and their ability to hear soft sounds ([Kochkin, 2005]); thus, the common report that people are either shouting or mumbling. Children with S/PHL also have reported that conversational speech is not loud enough after changing to a new prescription fit ([Gabbard et al, 2003]; [Scollie et al, 2010]). Importantly, dissatisfaction with HA loudness has been identified as a factor related to the ultimate fitting failure—that is, discontinuing HA use—especially in young adults with S/PHL who feel that HAs make background noise too loud or cause headaches ([Cameron et al, 2008]).

Forgoing aided loudness behavioral judgments can risk overlooking effects of loudness summation, channel summation, and binaural summation (for bilateral HA users) ([Mueller and Bentler, 2005]). The benefits and drawbacks of loudness normalization notwithstanding, unless directly measured, it cannot be ascertained to what extent an individual HA user is experiencing normalized aided loudness even when clinicians assure attainment of targets through real-ear probe-tube verification procedures ([Johnson and Cox, 2013]). Promoting aided loudness judgments in the verification phase of HA fitting is not a new consideration ([Mueller, 2003]), but even at the validation stage, there are barely half of practitioners confirming their fitting objectives with aided loudness judgments, few who validate with a self-assessment survey, and some who do not use prescription fit methods at all ([Tharpe et al, 2001]; [Mueller, 2003]; [Humes and Amos, 2009]; [American Speech-Language-Hearing Association, 2010]; [Kirkwood, 2010]; [Mueller and Picou, 2010]; [Mueller, 2014]). Whereas aided loudness perception has been studied in the larger population of listeners with mild-to-moderate hearing loss, there have been few reports that describe aided loudness outcomes in those with S/PHL, including to what degree they attain loudness normalization and whether it has desirable effects. The current study examined aided loudness growth in experienced young-adult HA users who had early-onset bilateral S/PHL. The purpose was to ascertain the frequency of aided loudness normalization and to characterize any instances of nonnormal aided loudness perception.

Early studies of the relationship between signal level and its subjective loudness provide a set of descriptors that might be useful toward characterizing the graphic patterns produced by the listener group we chose to study. It should be noted, however, that the current study was not concerned with the diagnostic significance of the patterns proposed in that early work or other purported implications about how auditory/neural systems functioned. Thus, reference to those patterns in the present report is not intended to suggest cause or locus of hearing loss, but merely to provoke the visual representation of a familiar function. Examples of patterns of loudness growth in the presence of hearing loss are available, first, in classic studies of recruitment. Recruitment patterns described by [Harris et al (1952)] included asymptotic, straight-line, delayed, and delayed plus asymptotic, all of which could be subsumed under the label “complete recruitment” ([Moore, 2007]), wherein high-level sounds are about as loud as in normal hearing (NH). In contrast to recruitment, [Davis and Goodman (1966)] and [Fowler (1965)] identified “decruitment” (slow growth, or less-rapid-than-normal growth of loudness with intensity), similar in shape to “underrecruitment” ([Moore, 2007]), where loudness remains lower than normal, and “partial recruitment” or “loudness reversal” ([Priede and Coles, 1974]). It is noted, however, that an explanation of decruitment as a loudness phenomenon has been disputed ([Mencher et al, 1973]). A third pattern of loudness growth in the literature is “overrecruitment” (or hyperacusis), with a recruitment curve that does not decelerate at high input intensity levels, resulting in high-level sounds becoming louder than normal ([Moore, 2007]).

In addition, [Marozeau and Florentine (2007)] named an “Attenuation” pattern, that runs parallel to the normal loudness function, as expected with conductive hearing loss; and a new concept called “Softness Imperception,” defined as an inability to hear soft sounds, with faster than normal growth at midlevels and near normal at high levels. The latter pattern is visually similar to a horizontally shifted version of Harris’s asymptotic recruitment. As the reality of Softness Imperception remains controversial ([Moore, 2004b]; [Palmer, 2013]) and concerns intensity levels very near threshold of detection that were not examined closely in the current study, the Softness Imperception pattern was not used here as a guide for categorizing our participants’ loudness growth curves.

[Villchur and Killion (2008)] called for documenting the prevalence of the loudness patterns described by Harris, suggesting that that knowledge could guide HA fits to restore an individual’s loudness growth curve to normal. Not predictable from overall satisfaction ratings or HA parameters, patterns of aided loudness growth must be directly observed, a procedure initiated in groups with lesser degrees of hearing loss ([Shi et al, 2007]; [Johnson and Cox, 2013]) but not yet in adults with S/PHL of early onset. In the current study of aided loudness growth, it was anticipated that (a) the classic recruitment pattern of steeper than normal loudness would predominate because of the nature and degree of hearing loss, (b) any instances of normalized loudness would be associated with better high-frequency detection thresholds and speech-recognition scores, and (c) reports of listening satisfaction would reflect clients’ long-term experience living with the limitations of their auditory system and HA technology and thus would be neither strongly positive nor strongly negative.

METHODS

Before collection of data, the design and methods of the research study described below were approved by the Institutional Review Board of the authors’ university.

Participants

Hearing Loss Group

Participants were a convenience sample of 23 university students with long-standing S/PHL who agreed to participate in this study for pay and met the criteria of early onset of hearing loss, initiation of HA use by age 3 yr, and a newly dispensed HA judged via electroacoustic measurement to be functioning according to specifications at the time of the study. There were 15 females and 8 males, 19–44 yr of age (mean = 23.7 yr). Client records indicated that all devices were initially adjusted with a given manufacturer’s fitting algorithm by one of five clinical audiologists specialized in fitting HAs for S/PHL. Clients were then allowed an opportunity to use the HA in real-world settings, after which adjustments were made if needed to customize the fit according to user feedback and expressed preferences, applying the principles of both “audiologist-driven” and “patient-driven” fitting strategies ([Palmer and Lindley, 2002]). Evaluation of the fit of the participants’ HAs or the fitting procedure was not part of the protocol of this study. Data collection described below occurred with the participants’ HAs as dispensed and worn at the time of testing, after approximately two weeks of use.

As shown in [Table 1], hearing loss was moderately severe for four participants and severe or profound for 19 participants (better-ear pure-tone average thresholds [PTA], mean = 94.2 dB hearing loss, standard deviation [SD] = 9.3). High-frequency average thresholds were severe–profound for all but S2 (73–117 dB hearing loss; mean = 96.8 dB hearing loss, SD = 15.2). Onset of hearing loss was by age 3, and HAs had been worn since an early age with educational support services continuing throughout schooling for all participants. Clinician’s choice of auditory speech-reception assessments focused on prospective recommendations for listening training and was guided by each listener’s performance on a four-choice spondee test or previous word-recognition test at a normal conversational-speech level standard for the clinic test environment. Central Institute for the Deaf (CID) W-22 word lists and/or CID everyday sentences (keyword scoring) were administered where there was a history of better than 0% correct on a word or sentence test or where the spondee response pattern indicated potential value in pursuing phonemic-cue based testing.

Table 1 Audiologic Characteristics of Participants with S/PHL

Participant ID	Better Ear	Threshold of Detection in dB Hearing Loss for Audiometric Test Frequencies (Hz)						PTA[†]	High- Freq Ave[‡]	Etiology	Age at Onset	Age at Test	Aided Speech Discrim[§]
		250	500	1k	2k	4k	8k
1	R	20	35	60	75	75	70	57	70	Unknown	Birth	19	Sp: 20/20
													W-22: 48%
2	R	40	60	60	60	65		60	62	Unknown	Birth	25	W22: 72%
3[*]	L	55	50	60	75	85	NR[‖]	62	73	Possible genetic	Birth	22	Sp: 20/20
4[*]	L	80	70	65	65	90		67	73	Meningitis	8 mo	20	Sp: 20/20
													W22: 0%
5	L	60	75	85	75	70	105	78	77	Unknown	3 yr	44	Sp: 20/20
													W22: 0%
6	L	60	75	90	80	80	65	82	83	Unknown	1 yr	20	W22: 24%
7	R	70	75	95	85	75	70	85	85	Kallmann’s	Birth	21	W22: 0%
8	R	60	70	90	105	NR	NR	88	103	Possible genetic	2 yr	19	CID: 82%
													W22: 0%
9	R	55	75	90	100	105	NR	88	98	MMR vaccine	18 mo	22	Sp: 20/20
													W22: 32%
10	L	75	90	100	80	60	75	90	80	Unknown	Birth?	28	Sp: 20/20
													W22: 48%
11	R	70	90	90	90	85	105	90	88	Possible genetic	3 yr	23	W22: 32%
12	R	75	80	100	NR	NR	NR	90	113	Unknown	Birth	24	Sp: 12/20
13	R	65	80	90	100	110		90	100	Genetic	Birth	24	Sp: 3/20
14[*]	R	90	80	95	100	110	NR	92	102	Unknown	Birth	20	Sp: 18/20
													W22: 0%
15	R	60	80	95	105	110	NR	93	103	Unknown	Birth	21	Sp: 20/20
													W22: 48%
16	R	80	85	100	100	105	100	95	102	Unknown	Birth	19	Sp: 18/20
													W22: 0%
17	L	80	85	100	110	NR	NR	98	108	Ototoxic drugs	3 yr	23	Sp: 1/20
18	L	90	90	100	110	115	NR	100	108	Waardenberg	Birth	21	CID: 19%
19	L	100	85	100	115			100	108	Unknown	Birth	23	Sp: 15/20
20	R	80	90	105	110	120	NR	102	112	Unknown	Birth	27	Sp: 18/20
													W22: 0%
21	R	85	100	110	115	NR	NR	108	113	Meningitis	3 yr	21	Sp: 2/20
22[*]	L	100	105	110	115			110	113	Unknown	Birth	24	Sp: 7/20
23	L	100	110	110	NR	NR	NR	112	117	Ototoxic drugs	Birth	35	Sp: 6/20

Notes: Empty cells = did not test.

^* Unilateral HA user.

^† PTA: Better-ear pure-tone-average threshold in dB hearing loss for 500, 1000, and 2000 Hz.

^‡ High-freq PTA: Better-ear pure-tone-average threshold in dB hearing loss for 1000, 2000, and 4000 Hz for bilateral HA fits. Aided-ear thresholds were averaged for unilateral fits (S3, S4, S14, and S22).

^§ Speech Discrim: Auditory-only speech-discrimination performance, aided; Sp = 4-choice spondee identification; number correct out of 20 trials; W22 = CID W-22 word list, percent of words correct out of 50; CID = percent of keywords correct in 10 CID Everyday Sentences.

^‖ NR = No response at the limits of the audiometer. Pure-tone averages substituted a value 5 dB above the limits of the audiometer.

Local Norms (NH) Group

Fifteen university students with NH from the same institution as the S/PHL group were recruited to generate local norms. They demonstrated normal otoscopy and tympanometry, pure-tone thresholds at 15 dB hearing loss or better for the octave frequencies 125–8000 Hz, negative otologic history, and no prior experience with loudness rating. Ten females and five males, 18–40 yr old (mean = 23.3 yr, SD = 7.0), volunteered to participate for pay.

Loudness Rating

Loudness Rating Conditions

Loudness perception was measured with broadband and narrowband signals, both speech and nonspeech: (a) 5-sec segments from the Connected Speech Test (CST) ([Cox et al, 1987]; [Cox and Gray, 2001]); (b) 5-sec duration warble tones (Wbls) with 5% warble, 5 Hz modulation rate, centered at 500, 1000, or 2000 Hz (referred here as Wbl5, Wbl1k, and Wbl2k, respectively); and (c) 5-sec one-third-octave band-filtered segments of the California Consonant Test (CCT; [Owens and Schubert, 1977]), male voice, with silent intervals removed between words, and filters centered at 500, 1000, or 2000 Hz (referred here as CCT5, CCT1k, and CCT2k, respectively). Aided loudness ratings for band-limited signals can be diagnostic in pinpointing frequency-specific problem areas and guiding subsequent HA adjustment for loudness comfort. The CCT has been useful in exploring bandwidth and frequency effects on speech understanding ([Schwartz and Surr, 1979]; [Prendergast, 2005]; [Lindley 2009]).

Loudness Rating Response Task

Loudness ratings were collected using the Contour Test of loudness perception ([Cox et al, 1997]). A printed copy of the loudness categories ([Table 2]) and instructions from [Cox et al (1997)] were provided for participants to read: “The purpose of this test is to find your judgments of the loudness of different sounds. You will hear sounds that increase and decrease in volume. You must make a judgment about how loud the sounds are. Pretend you are listening to the radio at that volume. How loud would it be? After each sound, tell me which of these categories best describes the loudness. Keep in mind that an uncomfortably loud sound is louder than you ever would choose on a radio no matter what mood you are in.” Instructions were clarified for the S/PHL group using sign language, as needed.

Table 2 Scale for Rating Loudness

Loudness Category Descriptor	Category Number
Nothing	0
Very quiet (very soft)	1
Quiet (soft)	2
Comfortable, but quiet (soft)	3
Comfortable	4
Comfortable/loud	5
Loud, but OK	6
Ouch (uncomfortable)	7

Testing occurred in a double-walled audiometric test booth. Listeners were positioned in the calibrated sound field with head 1 m from the loudspeaker at a 0-degree azimuth. Participants with NH were tested binaurally unaided. Participants with S/PHL were tested with both ears aided, HAs at use settings, except for the four unilateral HA users noted in [Table 1] who wore an ear plug in their unaided ear during testing. Participants rated the listening conditions over two sessions within one week’s time. Roughly half began with the CST condition, followed by CCT5, CCT1k, and CCT2k, and lastly Wbl5, Wbl1k, and Wbl2k. The other participants received Wbls first and the CST condition last. Signals were presented in an ascending manner, beginning at 5 dB above threshold to avoid issues of loudness summation near threshold and continuing until the listener responded that it was uncomfortably loud or the sound field limit of 85 dB hearing loss (97.5 dB SPL) was reached. Immediately after each stimulus, participants rated the loudness by speaking or signing a number from 0 to 7. Four ascending runs, with a stimulus increment of 5 dB, were administered in each condition. A fixed 5-dB increment has been found to produce the same results as random 2- to 5-dB increments ([Cox et al, 1997]). Analyses of the results were conducted using SAS 9.4.

Hearing Aid Rating

To gauge HA benefit and satisfaction, we accessed clinic results of the Client-Oriented Scale of Improvement (COSI; [Dillon et al, 1997]). Participants rated how satisfactorily the new HA met the priority needs they had identified. The COSI is said to be an indicator of the relative and absolute ease of communication with a new HA ([Dillon et al, 1997]) and how well it satisfies a listener’s real-life concerns ([Cox, 2005]). COSI results have correlated well with a global measure of HA benefit/satisfaction ([Dillon et al, 1997]) and have supported findings of a commonality between benefit and satisfaction specifically for HA users with moderately severe to severe hearing loss ([Emerson and Job, 2014]).

RESULTS

Results of the Contour Test

Local Norms

The median signal level associated with a given Loudness Category descriptor ([Table 2]) was calculated across the four runs for each NH listener. The mean of these values for the group is shown for each Loudness Category in [Figure 1] and [Table 3]. Normal loudness for the local test environment was defined as the range between the 10th and 90th percentiles of NH medians, excluding extremes of the distribution following [Cox et al (1997)]. The upper and lower bounds of our local norms (10th and 90th percentile values) are shown by the small circles in [Figure 1]. Perceived loudness increased systematically through Loudness Category 6 in each listening condition. As expected, broadband speech (CST) was perceived to be louder than filtered speech or Wbls and resulted in a more linear loudness function, especially in the comfort region, likely due to spectral loudness summation across critical bands ([Ricketts and Bentler, 1996]; [Brand and Hohmann, 2001]), lending additional validity to the norms.

Figure 1 Results of the Contour Test of loudness perception in 7 listening conditions for 15 listeners with NH in the sound field (NH: broken lines, indicating the region between the 10th and 90th percentiles of the distribution of individual medians calculated over four runs) and 23 listeners with S/PHL with best-fit HAs (hearing loss: solid lines, representing the mean of individual medians over 4 runs). CST = 5 sec segments of the CST. W5, W1k, and W2k = narrowband-filtered Wbls centered at 500, 1000, or 2000 Hz, respectively. CCT5, CCT1k, and CCT2k = 5 sec segments of narrowband-filtered words of the CCT centered at 500, 1000, or 2000 Hz, respectively.

Table 3 Results of Contour Test of Loudness Perception for Participants with NH: Mean of Individual Participant’s Median Signal Levels in dB Hearing Loss (SD) and Number of Respondents Associated with Each Loudness Category

Loudness Category Descriptor	Loudness Category Number		CCT[†]			Wbl
			Frequency Band (Hz)			Frequency Band (Hz)
		CST[*]	500	1000	2000	500	1000	2000
Very quiet (very soft)	1	22.3 (2.2)	27.3 (3.3)	26.2 (3.3)	26.5 (2.3)	24.6 (1.7)	26.0 (3.4)	25.2 (3.3)
Quiet (soft)	2	31.5 (2.8)	43.8 (4.9)	40.0 (5.2)	40.7 (3.2)	36.3 (5.2)	38.7 (5.8)	37.3 (5.9)
Comfortable, but quiet (soft)	3	37.5 (4.3)	55.2 (4.3)	49.5 (5.1)	51.8 (4.5)	45.5 (6.4)	48.2 (5.9)	47.5 (4.0)
Comfortable	4	47.8 (4.8)	63.3 (5.1)	59.7 (4.7)	60.8 (4.5)	55.7 (6.2)	58.2 (5.2)	56.8 (4.8)
Comfortable/ loud	5	57.8 (6.0)	71.0 (5.0)	69.3 (5.6)	70.3 (4.5)	66.1 (5.9)	67.0 (5.8)	65.0 (6.2)
Loud, but OK	6	68.2 (6.9)	78.3 (5.7)	76.8 (5.7)	78.0 (4.3)	75.0 (7.3)	75.0 (6.1)	72.7 (6.0)
Ouch (uncomfortable)	7	75.7 (5.9)	80.0 (5.3)	80.5 (4.7)	81.3 (5.3)	77.3 (5.1)	80.0 (5.8)	77.3 (5.2)
		14	8	11	12	10	13	14

Note: n = 15 for loudness categories 1–6 in all conditions except Wbl 500 Hz where n = 14. At Loudness Category 7, n is shown below the SD.

^* CST, broadband signal.

^† CCT, narrowband-filtered words.

Patterns of Aided Loudness for Participants with S/PHL

Median signal levels associated with each Loudness Category were calculated as for the NH listeners, and the means of these values are shown in [Table 4] and [Figure 1] (large circles). There was greater variability among participants with S/PHL compared with NH participants, especially at higher frequencies, consistent with a wider range of detection thresholds for participants with S/PHL, yielding progressively fewer responses beyond the comfort region. Compared with NH participants, participants with S/PHL needed a stronger signal for an equivalent loudness experience up through the comfort region except with the lower bands of tones and filtered words. Above the comfort region, the average response from participants with S/PHL was within local norms.

Table 4 Results of Contour Test of Loudness Perception for Participants with Hearing Loss: Mean of Individual Participants’ Median Signal Levels in dB Hearing Loss (SD) and Number of Respondents Associated with Each Loudness Category

Loudness Category Descriptor	Loudness Category Number		CCT[†]			Wbl
			Frequency Band (Hz)			Frequency Band (Hz)
		CST[*]	500	1000	2000	500	1000	2000
Very quiet (very soft)	1	33.4 (6.2)	38.1 (3.6)	37.0 (7.4)	42.6 (11.2)	37.9 (7.8)	41.3 (8.3)	52.1 (13.5)
		23	22	23	23	23	23	23
Quiet (soft)	2	42.2 (6.6)	48.3 (2.6)	45.3 (7.2)	51.3 (10.2)	47.1 (8.9)	50.0 (8.6)	58.5 (12.5)
		23	22	23	22	23	23	20
Comfortable, but quiet (soft)	3	51.3 (7.9)	58.0 (3.8)	54.1 (8.2)	60.6 (11.0)	55.8 (9.5)	57.5 (9.2)	65.9 (11.9)
		23	22	23	21	23	23	20
Comfortable	4	60.3 (9.9)	65.4 (1.6)	62.2 (10.0)	68.8 (9.1)	62.6 (8.9)	63.6 (9.5)	71.0 (10.8)
		23	20	22	20	22	22	18
Comfortable/loud	5	67.4 (9.9)	72.0 (1.6)	68.5 (9.9)	74.1 (7.6)	69.9 (9.2)	69.4 (9.7)	75.2 (7.6)
		21	19	21	16	20	21	12
Loud, but OK	6	72.3 (7.0)	76.8 (0.8)	74.1 (8.3)	77.7 (5.6)	77.1 (8.4)	73.5 (8.5)	75.8 (6.5)
		21	17	19	11	20	17	9
Ouch (uncomfortable)	7	81.8 (2.9)	82.3 (2.5)	81.5 (4.2)	83.8 (2.7)	81.8 (3.9)	82.0 (4.4)	82.5 (2.0)
		19	12	17	8	10	15	7

^* CST, broadband signal.

^† CCT, narrowband-filtered words.

Analysis focused on Loudness Categories 2 through 6 (following [Shi et al (2007)], for example), avoiding expected effects of faster loudness growth near threshold and excluding the discomfort level for which we had a reduced data set in a region that is less informative about loudness growth ([Hellman and Meiselman, 1993]; [Silva and Epstein, 2012]). Loudness Category was treated as an interval scale, with loudness responses plotted against median signal level in dB hearing loss. Three participants had insufficient hearing to respond in the Wbl2k condition and one in the CCT1k condition. On 21 other trials, loudness responses were provided for fewer than four Loudness Categories, inadequate for reliable analysis. In one additional case, a linear fit failed to reach our criterion of R ² = 0.8. With those 26 exclusions, there remained 135 S/PHL loudness functions for analysis.

Graphic pattern of loudness growth was determined by overlaying each S/PHL loudness function onto the region of local norms for a given condition and grouping functions with similar orientation together, guided by patterns described in the literature ([Harris et al, 1952]; [Fowler, 1965]; [Davis and Goodman, 1966]; [Cox et al, 1997]; [Marozeau and Florentine, 2007]; [Moore, 2007]). Groupings were adjusted until exclusive criteria for group assignment were achieved and functions could be grouped reliably across authors. [Table 5] lists the incidence of each of the six identified patterns, with examples in [Figures 2]–[4].

• Normal. There were 30% of the 135 S/PHL functions contained completely within normal bounds ([Figure 1]). All but two participants produced a normal pattern in at least one condition.

• Near-normal. Nine S/PHL functions (7%) produced by seven participants failed to meet the criterion for normal, but by <5 dB, most often exceeding the upper bound of the normal region ([Figure 2]).

• Shifted. There were 10% of S/PHL functions roughly parallel to the normal region and situated beyond the NH 90th percentile by as much as 20 dB ([Figure 2]). This pattern appeared most often in the CST condition and was graphically similar to Jerger’s “no recruitment” ([Jerger, 1962]) and the Attenuation pattern expected with conductive hearing loss ([Marozeau and Florentine, 2007]) although all hearing loss in this study was sensorineural with no measurable conductive component. There were 12 participants who produced 1–3 shifted patterns each.

• Steep. There were 38% of S/PHL functions that were labeled steep ([Figure 3]). These 52 patterns were beyond the NH 90th percentile by >5 dB at lower intensities and approached NH bounds at higher intensities, with 41 functions intersecting the upper bound of the normal region. Steep functions were produced in one or more conditions by 19 of the 23 participants, with most occurrences in the broadband-speech condition (most likely affected by the intensity variations in the waveform) and in Wbl conditions.

• Hypersensitive. There were 10% of S/PHL functions 5 dB or more below the NH 10th percentile for some or all of the loudness categories, most often beginning at the comfort region ([Figure 4]). This suggests hypersensitivity, whereby the listener associates a greater than normal loudness with a given signal level. Hypersensitive patterns were least common, unique to five participants. These patterns resemble examples of overrecruitment (hyperacusis; [Jerger, 1962]; [Priede and Coles, 1974]).

• Shallow. There were 7% of S/PHL functions, produced by eight participants, that lay beyond the NH 90th percentile for most loudness categories and increasingly diverged as signal level increased ([Figure 4]). These are graphically similar to examples of decruitment (or “underrecruitment”) described by [Davis and Goodman (1966)].

Table 5 Results of Graphic Pattern Analysis of Aided Loudness Functions for 23 Participants with Hearing Loss

	Number of Participants Producing a Given Loudness Pattern in Each Listening Condition[†]							Number of Pattern Occurrences Over All Conditions (% of Total)
Loudness pattern[*]	CST	W5	W1k	W2k	CCT5	CCT 1k	CCT 2k
Within normal bounds	3	5	5	3	8	12	5	41 (30.4)
Near-normal	1	1	1		3	1	2	9 (6.7)
Shifted right	7	1	3	1		1	1	14 (10.4)
Steep	12	13	9	8	3	3	4	52 (38.5)
Hypersensitive		1	2		3	3	1	10 (7.4)
Shallow		1	1	1	2	1	3	9 (6.7)
Total n	23	22	21	13	19	21	16	135

Note: Empty cells indicate no incidence of a particular pattern. Where total n is <23, the dynamic range of the missing participants encompassed fewer than four loudness levels, producing an insufficient number of data points for analysis.

^* Graphic pattern descriptors of the shape or location of the loudness growth function: Within normal bounds = completely inside normal bounds. Near-normal = <5 dB divergent from normal bounds. Shifted right = above (higher than) normal signal levels, with quasi-normal loudness growth. Steep = above (higher than) normal signal levels, with a sharp rise, approaching or intersecting normal region, suggesting under-compression. Hypersensitive = Shifted left, below (lower than) normal signal levels along some portion of the function suggesting hypersensitivity. Shallow = above (higher than) normal signal levels with shallower than normal slope.

^† Listening conditions: CST = 5 sec segments of the CST. W5, W1k, W2k = narrowband filtered Wbls centered at 500, 1000, or 2000 Hz, respectively. CCT5, CCT1k, CCT2k = 5 sec segments of narrowband-filtered words of the CCT centered at 500, 1000, or 2000 Hz, respectively.

Figure 2 Examples of aided loudness response patterns on the Contour Test of loudness perception produced by individuals with S/PHL (open circles), shown in relation to the local norms (solid lines). (A–E) Normal pattern = completely within the bounds of the local norms for all loudness categories. (F–I) Near-normal = within 5 dB of norms. (J, K) Shifted = outside, parallel to norms. For participant and listening condition, see [Figure 1] and [Table 1].

Figure 3 Examples of the Steep pattern (L–O) of aided loudness responses on the Contour Test of loudness perception produced by individuals with S/PHL (open circles), shown in relation to the local norms (solid lines). For participant and listening condition, see [Figure 1] and [Table 1].

Figure 4 Examples of the (P, Q) Hypersensitive and (R, S) Shallow patterns of aided loudness responses on the Contour Test of loudness perception produced by individuals with S/PHL (open circles), shown in relation to the local norms (solid lines). For participant and listening condition, see [Figure 1] and [Table 1].

Effect of Frequency and Type of Signal on Graphic Loudness Pattern

For further analysis, patterns that shared general features were grouped within signal type (broadband speech, band-limited speech, and Wbl). Normal and Near-normal patterns were combined, forming a group of 50 functions. The Shifted and Shallow patterns also were combined, representing 23 functions that were beyond the NH 90th percentile but were not steeper than normal. Hypersensitive patterns were excluded as being particular to only a few participants.

Illustrated in [Figure 5] are results by signal type, pooled across frequency for Wbls and CCT words. Frequency of signal (Wbls or band-limited words) was found not to be significantly related to the occurrence of any one of the primary loudness patterns (Normal/Near-normal, Steep, or Shifted/Shallow; post hoc χ² p > 0.05). Steep patterns were significantly more frequent with broadband speech (CST) and Wbls; Normal/Near-normal patterns were significantly more likely with CCT words (band-limited); and Shifted/Shallow patterns occurred at statistically similar rates across signal types (χ² (df = 4) = 22.4; p < 0.001).

Figure 5 Frequency of patterns of aided loudness growth on the Contour Test of loudness perception produced by individuals with S/PHL when listening to broadband speech (CST), Wbls (combined results for 500, 1000, and 2000 Hz), or narrowband-filtered words (CCT combined results for the 500, 1000, and 2000 Hz filter bands).

Listener Profiles Associated with Graphic Loudness Patterns

A Normal/Near-normal pattern was the predominant aided loudness outcome for four participants (S1, S2, S7, and S8; see [Table 1]). They all demonstrated moderate/severe hearing loss, and three of them had the best auditory speech-reception skill among the S/PHL group. Only one participant (S4), with an etiology of meningitis, failed to produce any Normal or Near-normal patterns. This participant’s auditory discrimination test scores indicated the ability to extract information from the time-intensity waveform of speech, but did not evidence phonemic perceptual skill. Five other participants (S9, S10, S16, S21, and S23) produced only one Normal/Near-normal loudness pattern. Most of this group also had poor aided speech-discrimination scores as well as high-frequency PTA of 102–117 dB hearing loss. On closer inspection, there were 379 trial-by-trial loudness judgments (54% of 701 trials with responses in the range from Loudness Category 2 to 6) that fell outside the norms, and the majority of these came from participants with poorer high-frequency hearing (>100 dB hearing loss).

A Steep pattern was the predominant outcome for only five subjects (S3, S13, S14, S21, and S22). Their commonality was little or no phonemic-cue-based perception. Although all participants were predicted to produce at least some indication of a Steep pattern, four participants with S/PHL produced no Steep patterns (S1, S4, S12, and S23); these participants had dissimilar audiological profiles except for congenital onset of hearing loss.

The Hypersensitive pattern was linked to one participant (S9) who produced half of the 10 occurrences. She had a severe hearing loss with some auditory word identification skill. The remaining Hypersensitive patterns came from four other participants (S4, S5, S11, and S17), all but one of whom had severe hearing loss like S9. This pattern is concerning, calling for additional testing and HA readjustment to confirm comfort and appropriate safeguards against noise-induced hearing loss ([Macrae, 1995]), as was provided to the five affected participants in the current study.

Shallow patterns, while among the least frequent, were produced once or twice by a third of the participant group (S1, S2, S4, S6, S13, S16, S17, and S19) and in every condition except broadband speech. Individuals with a Shallow pattern had diverse hearing and perceptual characteristics, including degree of hearing loss, auditory speech-recognition skill, and constellations of other aided loudness perception patterns; however, six of them would have been candidates for a cochlear implant. They chose to remain with acoustic stimulation to avoid the costs or risks associated with cochlear implant surgery and rehabilitation or because they did not deem the expected communication gains to be worthwhile. Shallow patterns in the context of diagnostic procedures would have identified decruitment as suggestive of an auditory system unable to receive, transmit, or interpret acoustic stimuli effectively ([Mencher et al, 1973]; [Stach, 1998]), a condition unlikely to change with HA technology. Such a pattern would signal consideration of HA modifications in the frequency range exhibiting the furthest departure from the local loudness norms.

Stability of Contour Test Responses over Runs

To assess stability of the responses on the Contour Test from participants with S/PHL, we examined how often they responded with the same Loudness Category at a given intensity level from run to run. All 644 runs were included, even those not useable for pattern analysis. Consistency of loudness ratings (intrasubject differences collapsed across conditions) is shown in [Table 6]. From run to run, participants with S/PHL chose the same Loudness Category for a given signal level 88–89% of the time. When Loudness-Category response did differ on a subsequent run, participants tended to select the adjacent loudness descriptor, softer after run 1 and louder after runs 2 and 3. Differences of >1 Loudness Category from run to run occurred only 7 times in over 5,300 run comparisons. These instances all occurred in broadband or low-pass conditions (CST, CCT5, or Wbl5).

Table 6 Consistency of Loudness Ratings from Run to Run, Measured in Percent of Instances in Which a Rating at a Given Signal Level was the Same, Softer, or Louder

	Percent of Instances of Differences in Run-to-Run Responses
Degree of run-to-run difference[*]	Run 2 vs. 1	Run 3 vs. 2	Run 4 vs. 3
Softer by 3	0.1
Softer by 2		0.1
Softer by 1	8.3	4.5	4.5
No difference	88.2	88.8	89.0
Louder by 1	3.3	6.6	6.4
Louder by 2	0.1		0.1
Louder by 3		0.1

Note: Empty cells have a frequency of zero.

^* Number of run-to-run comparisons was 1769–1778.

R ² was calculated to assess improvement in fit of loudness growth function from one run to the next. Fit was good on 97% of the 644 runs, regardless of run number (R ² ≥ 0.90 on 600 runs and ≥0.80 on 26 additional runs). In nearly all instances, R ² showed no substantial change with increasing number of runs. If a participant’s first run did not produce a function with R ² of at least 0.80, additional runs did not improve fit. In the 26 instances of R ² between 0.80 and 0.90, most occurred in a 1000 or 2000 Hz band-limited condition, and almost all were produced by the same six participants (S1, S6, S12, S13, S19, and S23). The remaining 18 runs had R ² values below 0.80 and occurred in the high-frequency conditions, CCT2k and Wbl2k. These less reliable outcomes were associated with three participants (S16, S19, and S23) who had profound hearing loss and demonstrated marginal benefit from auditory speech cues. The 135 functions discussed in the foregoing sections on patterns of loudness growth all had runs with R ² >0.80.

Slope Analysis of Aided Loudness Perceptions

As with pattern analysis, slope analysis focused on the range from Quiet-Soft to Loud-but-OK. Loudness responses were converted to a numeric scale (see [Table 2]), and a least-squares linear regression analysis model was employed to fit median signal level to loudness response, including only instances in which listeners responded at at least four signal levels. The linear regression fits were good in most cases, with R ² ≥ 0.900 for 97.8% of the S/PHL curves and 97.1% of the NH curves. One S/PHL function was omitted from slope analyses because of poor fit (R ² = 0.705 in the CCT5 condition for S20). For the 135 acceptable S/PHL functions with four or more points, R ² ranged from 0.845 to 1.0 (mean = 0.978), and for the 105 NH functions, R ² ranged from 0.810 to 1.0 (mean = 0.980).

A normalized aided loudness slope was defined as any value lying between the 10th and 90th percentiles of the NH distribution of slopes for a given listening condition. Results are summarized in [Figure 6] and [Table 7]. S/PHL functions were steeper on average than NH functions, with more instances of extremely high values and a wider range of slopes, although there was also one NH outlier in several conditions. Overall, 52% of the 135 S/PHL functions had slopes within normal bounds, 38% were steeper (above the 90th percentile), and 10% were shallower (below the 10th percentile). Differences between the NH and S/PHL distributions were statistically significant in the CST, CCT1k, Wbl5, and Wbl1k conditions (t = 2.17–3.23 (df for unequal variances = 33.3 to 36.0); p < 0.01 for Wbl1k; p < 0.05 for CST, Wbl5, and CCT1k), although it is important to note that at least half of the S/PHL slopes in those conditions were within the norms. In addition, whereas differences between listener groups were not statistically significant in the other three conditions (CCT5, CCT2k, and Wbl2k; p > 0.05), the S/PHL group produced normalized slopes only about a third of the time in those conditions.

Figure 6 Distribution of slopes of loudness functions produced by aided listeners with S/PHL (hearing loss, filled circles), and listeners with NH, in seven listening conditions (see [Figure 1]) on the Contour Test of loudness perception. Random horizontal jitter added for visibility of overlaid points with same y value. Each box = 25th–75th percentile. Whiskers = 1.5 times the interquartile range. Diamonds = mean and upper/lower 95% confidence points. Horizontal line across both distributions = grand mean for the condition. Asterisk = significant difference between hearing loss and NH distributions for a given condition.

Table 7 Mean and SD of Slopes of Loudness Functions (Linear Regression) for Each Listening Condition for Participants with NH and with Hearing Loss and Local Norms Defined as the Values at the 10th and 90th Percentiles of the Distribution of NH Slopes

Signal	NH Mean (SD)	Hearing Loss Mean (SD)	Local Norms 10th and 90th Percentiles	Percent of Hearing Loss Slopes Within Local Norms, %
CST	0.108 (0.025)	0.131 (0.040)	0.083–0.144	61
CCT 500 Hz	0.117 (0.020)	0.132 (0.044)	0.095–0.140	32
CCT 1 kHz	0.111 (0.024)	0.132 (0.035)	0.085–0.139	67
CCT 2 kHz	0.107 (0.015)	0.118 (0.035)	0.094–0.127	31
Warble 500 Hz	0.106 (0.028)	0.138 (0.050)	0.080–0.131	64
Warble 1 kHz	0.113 (0.027)	0.151 (0.044)	0.078–0.140	57
Warble 2 kHz	0.118 (0.031)	0.142 (0.038)	0.083–0.143	38

Note: See [Table 5] for description of listening conditions.

Normalized Slope Profile

No discernible profile was apparent between available audiological characteristics of the participants and numeric slope of their loudness functions. When slopes were coded as “In” or “Out” of the region of local norms and the data were pooled across listening conditions, a profound hearing loss was associated with the likelihood of a nonnormal (“Out”) slope (χ² (df = 1) = 4.63, p < 0.05) as shown in [Figure 7], top panel. The pattern of results for speech-reception skill was more complex ([Fig. 7], bottom panel). For analysis, auditory speech reception for W-22 word lists and/or CID everyday sentences, keyword scoring, was coded as follows: (a) Low, 0%; (b) Moderate, <40%; or (c) High, 40% or higher. Neither extreme of speech-reception skill was reliably predictive, but an intermediate level of speech-reception skill tended to be associated with slopes that were outside the norms (χ ² (df = 2) = 7.25, p < 0.05). These results are consistent with the dichotomy proposed by [Vermeulen et al (1995)] and applied to HA fitting in young children by [Brokx et al (1997)]; namely, that listeners with PTA <90 dB typically will evidence good phonemic-cue-based auditory speech perception with HAs, and those with PTA >110 dB will likely perceive only what can be distinguished on the basis of intensity changes over time. In particular, the current finding fits with the “hearer” versus “feeler” versus mixed-group distinction made by [Cramer and Erber (1974)]. Some of our listeners were neither fully like “feelers” who rely primarily on time-intensity patterns for speech understanding and do not experience normal aided loudness growth nor were they clearly “hearers” who can extract some phonemic features from the auditory signal and attain normalized aided loudness in some conditions.

Figure 7 Percent of aided loudness functions that had slopes within (“In”) or outside of (“Out”) local norms (combined across all listening conditions) produced by listeners with S/PHL. Mod/Sev = moderate to severe.

Connections between Graphic Pattern and Slope of Loudness Growth

As shown in [Figure 8], there was a systematic relationship between slope and pattern of loudness growth for the S/PHL listeners (General Linear Modeling Analysis of Variance of Slope × 6 Graphic Patterns; F = 26.45; df = 5; p < 0.0001). Slopes for functions with Normal/Near-normal patterns overlapped Shifted pattern slopes, all of which were significantly lower than slopes from Steep or Hypersensitive patterns. These latter two patterns had the steepest average slopes compared with all other patterns. Shallow pattern slopes, though lowest on average, were not significantly different from Shifted pattern slopes.

Figure 8 Range of slope of aided loudness functions produced by listeners with S/PHL, grouped by pattern of loudness growth (1 = within Normal bounds; 2 = Near-normal, within 5 dB; 3 = Shifted right, but parallel to Normal; 4 = Recruitment-like/Steep; 5 = Hypersensitive; 6 = Decruitment-like/Shallow). Random horizontal jitter added for visibility of overlaid points with same y value. Each box = 25th–75th percentile. Whiskers = 1.5 times the interquartile range. Line across each box = mean slope for that pattern. Dotted line across the display = grand mean for all patterns.

Most often, Normal/Near-normal and Shifted patterns had normalized slopes, Steep and Hypersensitive patterns had slopes above normal bounds, and Shallow patterns had slopes below normal bounds (Logistic Regression analysis of Slope Pattern [In, Above, or Below norms] × 6 Graphic Patterns; Wald χ ² for the global null hypothesis [df = 5] = 37.51; p < 0.0001).

Loudness and Satisfaction Responses

All S/PHL listeners listed at least two priority needs on the COSI, and some as many as five needs. To summarize across all participants, only the top two priorities were considered for this report, focusing on the item that asked about satisfactory hearing with the new HA. Most of the participants (n = 19) reported that they heard satisfactorily in their top-rated need situations most of the time or almost always, but four participants (S6, S9, S11, and S18) responded that they heard satisfactorily only half of the time. All these least-satisfied participants demonstrated phonemic-cue-based auditory speech skills and produced one or two Normal/Near-normal loudness functions. Two of them (S9 and S11), however, indicated loudness tolerance problems on the Contour Test, producing 1–5 Hypersensitive patterns each.

DISCUSSION

The intent of the present study was to describe the experience of aided loudness perception in young-adult HA users with early-onset S/PHL. Notably, the results demonstrated that a recruitment-like pattern cannot be the assumed outcome for these listeners and that each individual will display a unique profile of normal aided loudness perception, along with nonnormal patterns that might indicate a need for further intervention. Of particular clinical significance are the Hypersensitive (hyperacusis) and Shallow (slow growth) patterns, as also seen in [Geller and Margolis (1984)], [Margolis (1985)], and [Shi et al (2007)], and perhaps representative of outliers identified as “sound sensitive” or “sound addicts” by [Elberling (1999)]. It was also observed that listeners in the present study expressed satisfaction with their overall listening experience, even given their unfavorable measured loudness outcomes. The size of our participant group, as well as available client and HA information, limited interpretation of the results. Validating the loudness growth patterns that emerged will require a larger sample of HA users with early-onset S/PHL. A laboratory research procedure is also recommended, not bound by the constraints of a typical clinic appointment, that can require use of the same HA technology across the study group, a uniform protocol administered within an extended timeframe, and full compliance throughout the study including follow-up interventions.

We do not yet know the degree to which loudness for S/PHL listeners can be normalized, given available high-power HA technology and an impaired auditory system ([Kuk and Ludvigsen, 2000]). An individual with a history of long-term daily use of high-gain linear amplification with peak clipping, and limited potential for auditory-only speech recognition, may establish a preference for nonnormal loudness relationships and perception of speech via patterns in the time-intensity waveform as an aid to lipreading ([Flynn et al, 1998]; [Kuk et al, 2012]). Accustomed to distortion effects on important amplitude and voicing cues, these HA users often dislike the signal delivered by current-day technologies. Thus, they may shift the fitting toward settings that reproduce the more familiar nonnormal aided loudness perceptions beneficial to lipreaders. If this occurred during the HA fittings with our participants, it may partly explain their aided listening satisfaction. Frequency-gain prescriptions for S/PHL ([Byrne et al, 1990]; [Gabbard et al, 2003]; [Ching et al, 2015]) have not received as much attention as for lesser degrees of hearing loss although various configurations of power compression HAs for this group have been evaluated or recommended ([Faulkner et al, 1992]; [Barker et al, 2001]; [Flynn and Schmidtke, 2002]; [Condie and Tchorz, 2004]; [Marriage et al, 2005]; [Beck and Schum, 2006]; [Sockalingam et al, 2011]). Results from trying to transition S/PHL HA users to prescription targets and “maturation of benefit” have been uneven ([Cox, 1993]; [Cox et al, 2007]; [Scollie et al, 2010]; [Convery and Keidser, 2011]). Postfitting interventions that have been effective for S/PHL listeners in changing even intensely negative reactions to amplification include (a) frequent communication and personal contact with the audiologist, (b) discussion of what the HA user should expect from newer technology, and (c) reflections on their listening experiences in a written log during adaptation to the new HA ([Gottermeier et al, 2016]). When clients learn that their first reactions are not atypical and are expected to change, they become more willing to allow time for acclimatization. For those who do perceive phonemic distinctions and for whom nonnormal loudness would be an obstacle to auditory speech recognition, there have also been some successes. [Kuk et al (2003)] found that speech recognition scores for their S/PHL group improved just one month after moving to a nonlinear HA.

Active listening in the form of auditory training may be worth exploring to spur acclimatization and, perhaps, modify aided loudness perceptions; however, it remains open to investigation whether general listening practice or working to optimize specific speech-perception skills can alter loudness judgments in S/PHL HA users. At-home auditory training practice is now easier because of the proliferation of “listening therapy” applications ([Childress, 2013], [2016]) for clients who are unable or unwilling to adhere to a regimen of repeated clinic visits. Remotely delivered encouragement and instructional supports from the clinician may be needed to ensure persistence with the at-home technology in the absence of face-to-face contacts. An alternative to active-listening training is to program the HA, itself, to help the listener adjust to changes in amplification. Individuals might listen in their everyday home and work sound environments while toggling between a program that mimics what they were used to and a program with the target prescription. Because tendencies may be to favor the familiar, clinicians could select an HA from a manufacturer with fitting software that incorporates automated acclimatization steps ([Mueller and Powers, 2001]) or a trainable HA with a “gain-learning” feature ([Palmer, 2012]; [2013]).

S/PHL listeners in the authors’ clinical experiences are sophisticated consumers, many of whom prefer HAs that enable them to control what they hear. We have been reminded that “the patient is not the expert” ([Palmer, 2013]), but individuals with early-onset S/PHL have diverse reasons why they use amplification and are not all alike in the value they place on spoken-language communication (for a discussion of the social, linguistic, and cultural characteristics of individuals who identify as Deaf, see [Padden and Humphries, 1990]; [2005]; and [Hyde and Power, 2006]). An in-depth inquiry on listener satisfaction, preferably a guided interview to clarify and expand on questionnaire responses, may aid in furthering our understanding of the practical importance of a given prescription fit or normalized loudness adjustments as they affect client-defined benefit and satisfaction. Whether modifications after the initial HA fit are controlled by the HA, the HA user, or the clinician, it will be advisable to monitor impact on listening satisfaction and HA usage in addition to loudness perceptions, aided versus unaided lipreading, and auditory speech recognition. Phonemic analysis of perceptual errors during an acclimatization process and probes of loudness judgments may justify continued efforts toward acceptance of a new (normal loudness) prescription fit or an agreement to return to a previous fit that supports primarily a visual speech-reception process (see [Barker et al, 2001]; [Souza et al, 2005]; [Beck and Schum, 2006]; [Davidson and Skinner, 2006]; [Davies-Venn et al, 2009]). When these decisions align with client motivations for using amplification, the impact on HA benefit and satisfaction may more likely be positive.

Given that the need to achieve normalized loudness remains equivocal and that adding loudness testing to the clinical protocol for listeners with S/PHL has been questioned at least since [Elberling (1999)], clinicians may deem obtaining aided loudness judgments to be a secondary goal relative to ensuring audibility, comfort, and speech intelligibility ([Moore, 2003]; but see also [Blamey and Martin, 2009]). Moreover, with prescription HA fitting today, clinicians might assume that loudness across the aided frequency range has already been accounted for, perhaps explaining why many choose not to measure loudness perceptions. Nevertheless, anomalies in aided loudness as seen in this study, even after application of best-fit practices, highlight the need to assess aided loudness growth within the dynamic range ([Shi et al, 2007]; [Palmer, 2013]). Whereas loudness ratings do not usually feed directly into a prescription for adjusting HA settings, they are expected to guide readjustment after fitting and verification ([American Speech-Language-Hearing Association, 2006]), especially in cases of S/PHL. With most proprietary fitting software, however, the audiologist cannot always predict the actual acoustic change during HA readjustment without also performing electroacoustic test measurements of the HA output. For example, when a client with S/PHL does not like the sound produced by the new HA and can express the nature of the problem (e.g., “hearing lots of unimportant sounds–too loud”), the most efficient initial approach may be to try the manufacturer’s “fitting tool” or “fitting assistant,” although it may not be evident specifically how the tool will modify the signal acoustically. The audiologist can be more confident in predicting the actual acoustic change during HA readjustment when the HA allows more direct control of type of compression (e.g., “semilinear”), compression ratio, ambient noise, or maximum power (dB SPL) output, or if the HA has a transparent frequency transposition feature. The actual acoustic change is less predictable with features such as “brightness,” “fullness,” “spatial sense,” “acclimation,” or “preference control.”

Experienced clinicians often are guided by previous fits made with a given model of HA and how it has responded to various settings. When speech-mapping results are still below recommended targets after readjustment, as frequently occurs with high detection thresholds, the remaining strategy for verification and validation is to return to the booth and fall back on measuring aided detection thresholds, MCL, and UCL, and obtaining subjective loudness judgments via the COSI or other self-report instrument. In instances of persistent problems, loudness rating with narrowband stimuli assists in determining if a particular frequency region needs further attention. It is acknowledged, however, that even the audiologist’s best efforts can be thwarted when the HA user accesses a manufacturer’s application to modify settings (within the boundaries established by the audiologist) after leaving the HA fitting.

The Contour Test appears to be useful for obtaining aided loudness ratings. Degree of variability in responses from the 15 NH listeners in the current study was generally smaller than that reported for the 45 individuals tested by [Cox et al (1997)] who exhibited greater variability at higher signal levels. There have not been consistent findings regarding between-subject variability for listeners with NH ([Rasmussen et al, 1998]; [Sherlock and Formby, 2005]), leading us to question whether the middle 80% of the NH distribution is an appropriate criterion for normal. Including borderline responses (up to 5 dB outside the norms) doubled the number of S/PHL listeners who demonstrated at least one (quasi)-normalized loudness function. The most critical components of the clinical implementation of loudness rating would be standard written instructions (including a labeled response scale), consistent methodology ([Jenstad et al, 1997]), and local norms collected in the same test environment, along with a determination of how strictly to define normal.

The present findings on patterns of aided loudness perception underscore the diversity of abilities, needs, and preferences that characterize the early-onset S/PHL population. Aided loudness outcomes subsequent to verification of prescription targets in the HA fitting process potentially have implications for further HA adjustments on the part of the clinician and possibly within the client over time. The results are also consistent with suggestions already in the literature that HA technology and fitting approaches for S/PHL do not yet fully address the unique needs and abilities of all those with S/PHL.

Abbreviations

CCT: California Consonant Test

CCT1k: California Consonant Test words, filtered at 1 kHz (narrowband-speech condition for loudness rating)

CCT2k: California Consonant Test words

filtered at 2 kHz (narrowband-speech condition for loudness rating): CCT5

California Consonant Test words: filtered at 500 Hz (narrowband-speech condition for loudness rating)

CID: CID Everyday Sentence Test

COSI: Client-Oriented Scale of Improvement

CST: Connected Speech Test (broadband-speech condition for loudness rating)

HA: hearing aid

NH: normal hearing

PTA: pure-tone average hearing threshold

SD: standard deviation

S/PHL: severe–profound bilateral hearing loss

Wbl1k: 1-kHz warble tone (narrowband nonspeech condition for loudness rating)

Wbl2k: 2-kHz warble tone (narrowband nonspeech condition for loudness rating)

Wbl5: 500-Hz warble tone (narrowband nonspeech condition for loudness rating)

Wbl: warble tone

• REFERENCES

• American Speech-Language-Hearing Association 2006 Preferred practice patterns for the profession of audiology [preferred practice patterns]. Available from www.asha.org/policy . Accessed 12 October 2016
• American Speech-Language-Hearing Association 2010. 2010 Audiology Survey Summary Report: Number and Type of Responses. Rockville, MD: American Speech-Language-Hearing Association;
• Baer T, Moore BC, Kluk K. 2002; Effects of low pass filtering on the intelligibility of speech in noise for people with and without dead regions at high frequencies. J Acoust Soc Am 112 3 Pt. 1 1133-1144
• Barker C, Dillon H, Newall P. 2001; Fitting low ratio compression to people with severe and profound hearing losses. Ear Hear 22 (02) 130-141
• Beck D, Schum D. 2006 Amplification for people with severe and profound hearing loss. News from Oticon: Audiological Research Documentation. Retrieved from http://www.pro.oticonusa.com/∼asset/cache.ashx?id=20401andtype=14andformat=web . Accessed 7 March 2012
• Bentler RA, Cooley LJ. 2001; An examination of several characteristics that affect the prediction of OSPL90 in hearing aids. Ear Hear 22 (01) 58-64
• Blamey PJ, Martin LF. 2009; Loudness and satisfaction ratings for hearing aid users. J Am Acad Audiol 20 (04) 272-282
• Boothroyd A. 1984; Auditory perception of speech contrasts by subjects with sensorineural hearing loss. J Speech Hear Res 27 (01) 134-144
• Brand T, Hohmann V. 2001; Effect of hearing loss, centre frequency, and bandwidth on the shape of loudness functions in categorical loudness scaling. Audiology 40 (02) 92-103
• Brokx JP, Snik AF, Pepers-van Lith AI. 1997; Fitting hearing aids in children with severe hearing loss. Scand Audiol Suppl 46: 38-42
• Byrne D. 1996; Hearing aid selection for the 1990s: where to?. J Am Acad Audiol 7 (06) 377-395
• Byrne D, Parkinson A, Newall P. 1990; Hearing aid gain and frequency response requirements for the severely/profoundly hearing impaired. Ear Hear 11 (01) 40-49
• Cameron B, Cunningham E, Lindner A, Nicol L, Chenoweth L, Driscoll C. 2008; Hearing aid use and satisfaction in young Australian adults with severe to profound hearing loss. Aust N Z J Audiol 30: 59-72
• Childress T. 2013 Perusing, choosing and using apps. Retrieved from http://bit.ly/Apps4HL . Accessed 17 February 2017
• Childress T. 2016 Apps for kids (and adults) with hearing loss: listening therapy. Retrieved from https://docs.google.com/spreadsheets/d/1jBcMgGGKgUY2myJz8749IIZIWFwjlxBUjcy35SnahPo/edit#gid=9 . Accessed 17 February 2017
• Ching TY, Hill M, Dillon H. 2008; Effect of variations in hearing-aid frequency response on real-life functional performance of children with severe or profound hearing loss. Int J Audiol 47 (08) 461-475
• Ching TY, Johnson EE, Hou S, Dillon H, Zhang V, Burns L, van Buynder P, Wong A, Flynn C. 2013; A comparison of NAL and DSL prescriptive methods for paediatric hearing-aid fitting: predicted speech intelligibility and loudness. Int J Audiol 52 (Suppl. 2) S29-S38
• Ching TY, Quar TK, Johnson EE, Newall P, Sharma M. 2015; Comparing NAL-NL1 and DSL v5 in hearing aids fit to children with severe or profound hearing loss: goodness of fit-to-targets, impacts on predicted loudness and speech intelligibility. J Am Acad Audiol 26 (03) 260-274
• Condie R, Tchorz J. 2004 High-power hearing aids and children. Retrieved from http://www.audiologyonline.com/articles/pf_article_detail.asp?article_id=576 . Accessed 14 April 2011
• Convery E, Keidser G. 2011; Transitioning hearing aid users with severe and profound loss to a new gain/frequency response: benefit, perception, and acceptance. J Am Acad Audiol 22 (03) 168-180
• Cox R. 2005; Choosing a self-report measure for hearing aid fitting outcomes. Semin Hear 26: 149-156
• Cox RM. 1993; On the evaluation of a new generation of hearing aids. J Rehabil Res Dev 30 (03) 297-304
• Cox RM, Alexander GC, Gilmore C. 1987; Development of the connected speech test (CST). Ear Hear 8 (05) (Suppl.) 119S-126S
• Cox RM, Alexander GC, Taylor IM, Gray GA. 1997; The contour test of loudness perception. Ear Hear 18 (05) 388-400
• Cox RM, Alexander GC, Gray GA. 2007; Personality, hearing problems, and amplification characteristics: contributions to self-report hearing aid outcomes. Ear Hear 28 (02) 141-162
• Cox RM, Alexander GC, Johnson J, Rivera I. 2011; Cochlear dead regions in typical hearing aid candidates: prevalence and implications for use of high-frequency speech cues. Ear Hear 32 (03) 339-348
• Cox RM, Gray GA. 2001; Verifying loudness perception after hearing aid fitting. Am J Audiol 10 (02) 91-98
• Cramer KD, Erber NP. 1974; A spondee recognition test for young hearing-impaired children. J Speech Hear Disord 39 (03) 304-311
• Davidson LS, Skinner MW. 2006; Audibility and speech perception of children using wide dynamic range compression hearing AIDS. Am J Audiol 15 (02) 141-153
• Davies-Venn E, Souza P, Brennan M, Stecker GC. 2009; Effects of audibility and multichannel wide dynamic range compression on consonant recognition for listeners with severe hearing loss. Ear Hear 30 (05) 494-504
• Davis H, Goodman AC. 1966; Subtractive hearing loss, loudness recruitment and decruitment. Ann Otol Rhinol Laryngol 75 (01) 87-94
• Dillon H, James A, Ginis J. 1997; Client Oriented Scale of Improvement (COSI) and its relationship to several other measures of benefit and satisfaction provided by hearing aids. J Am Acad Audiol 8 (01) 27-43
• Elberling C. 1999; Loudness scaling revisited. J Am Acad Audiol 10 (05) 248-260
• Emerson L, Job A. 2014; Use of the client oriented scale of improvement (COSI) and international outcome inventory of hearing aids (IOI-HA) as a clinical outcome measure in a rural community. Egypt J Ear Nose Throat Allied Sci 15: 225-230
• Erber NP. 1979; Speech perception by profoundly hearing-impaired children. J Speech Hear Disord 44 (03) 255-270
• Faulkner A, Ball V, Rosen S, Moore BC, Fourcin A. 1992; Speech pattern hearing aids for the profoundly hearing impaired: speech perception and auditory abilities. J Acoust Soc Am 91 4 Pt. 1 2136-2155
• Faulkner A, Rosen S, Moore BC. 1990; Residual frequency selectivity in the profoundly hearing-impaired listener. Br J Audiol 24 (06) 381-392
• Flynn M, Schmidtke M. 2002 Four fitting issues for severe or profound hearing impairment. Hear Rev. Retrieved from http://www.hearingreview.com/2002/11/four-fitting-issues-for-severe-or-profound-hearing-impairment/ . Accessed 14 April 2011
• Flynn MC, Dowell RC, Clark GM. 1998; Aided speech recognition abilities of adults with a severe or severe-to-profound hearing loss. J Speech Lang Hear Res 41 (02) 285-299
• Fowler EP. 1965; Some attributes of “loudness recruitment” and “loudness decruitment”. Ann Otol Rhinol Laryngol 74: 500-506
• Gabbard S, O’Grady G, Kuk F. 2003; Fitting digital non-linear aids on children with severe-to-profound losses. Hear Rev 10 (07) 44, 46-49
• Gale E. 2011; Exploring perspectives on cochlear implants and language acquisition within the deaf community. J Deaf Stud Deaf Educ 16 (01) 121-139
• Geers AE, Moog JS. 1992; Speech perception and production skills of students with impaired hearing from oral and total communication education settings. J Speech Hear Res 35 (06) 1384-1393
• Geller D, Margolis RH. 1984; Magnitude estimation of loudness. I: application to hearing aid selection. J Speech Hear Res 27 (01) 20-27
• Gottermeier L, De Filippo C, Clark C. 2016; Trials of a contralateral hearing aid after long-term unilateral cochlear implant use in early-onset deafness. Am J Audiol 25 (02) 85-99
• Harris JD, Haines HL, Myers CK. 1952; Loudness perception for pure tones and for speech. AMA Arch Otolaryngol 55 (02) 107-133
• Hellman RP, Meiselman CH. 1993; Rate of loudness growth for pure tones in normal and impaired hearing. J Acoust Soc Am 93 (02) 966-975
• Humes L. 1982. Spectral and temporal resolution by the hearing impaired. In: Studebaker G, Bess F. The Vanderbilt Hearing-Aid Report: State of the Art–Research Needs. Upper Darby, PA: Monographs in Contemporary Audiology; 16-31
• Humes L, Amos N. 2009 , September 01 Are your hearing aid fittings “on target”?. The ASHA Leader
• Hyde M, Power D. 2006; Some ethical dimensions of cochlear implantation for deaf children and their families. J Deaf Stud Deaf Educ 11 (01) 102-111
• Jenstad LM, Cornelisse LE, Seewald RC. 1997; Effects of test procedure on individual loudness functions. Ear Hear 18 (05) 401-408
• Jerger J. 1962; Hearing tests in otologic diagnosis. ASHA 4: 139-145
• Johnson E. 2012 20Q: Same or different – comparing the latest NAL and DSL prescriptive tartets. Audiol Online. Retrieved from: http://www.audiologyonline.com/articles/20q-same-or-different-comparing-769 . Accessed 9 September 2016
• Johnson EE. 2013; Modern prescription theory and application: realistic expectations for speech recognition with hearing AIDS. Trends Amplif 17 (03) 143-170
• Johnson J, Cox R. 2013 Is normal loudness the appropriate goal for hearing aid fittings? Presented at the Annual Conference of the American Auditory Society, Scottsdale, AZ
• Keidser G, Dillon H, Dyrlund O, Carter L, Hartley D. 2007; Preferred low- and high-frequency compression ratios with moderately severe to profound hearing loss. J Am Acad Audiol 28: 17-33
• Keidser G, Hartley D, Carter L. 2008; Long-term usage of modern signal processing by listeners with severe or profound hearing loss: a retrospective survey. Am J Audiol 17 (02) 136-146
• Kirkwood D. 2010; Survey probes dispensers’ views on key issues raised by consumer reports. Hear J 63 (05) 17-26
• Kishon-Rabin L, Segal O, Algom D. 2009; Associations and dissociations between psychoacoustic abilities and speech perception in adolescents with severe-to-profound hearing loss. J Speech Lang Hear Res 52 (04) 956-972
• Kochkin S. 2005; MarkeTrak VII: customer satisfaction with hearing instruments in the digital age. Hear J 58: 30-43
• Kuk F, Andersen H, Baekgaard L. 2012; Hearing aids for severe-to-profound losses: business as usual?. Hear Rev 19 (03) 38-48
• Kuk F, Ludvigsen C. 2000; Hearing aid design and fitting solutions for persons with severe-to-profound losses. Hear J 53 (08) 29-37
• Kuk FK, Potts L, Valente M, Lee L, Picirrillo J. 2003; Evidence of acclimatization in persons with severe-to-profound hearing loss. J Am Acad Audiol 14 (02) 84-99
• Launer S, Kühnel V. 2001. Signal processing for severe-to-profound hearing loss. In: Seewald R, Gravel J. A Sound Foundation Through Early Amplification 2001: Proceedings of the Second International Conference. Stäfa, Switzerland: Phonak AG; 113-120
• Lindley G. 2009; Children seen to gain extra benefit from greater bandwidth, binaural compression. Hear J 62: 28-33
• Macrae JH. 1995; Temporary and permanent threshold shift caused by hearing aid use. J Speech Hear Res 38 (04) 949-959
• Malicka AN, Munro KJ, Baker RJ. 2010; Diagnosing cochlear dead regions in children. Ear Hear 31 (02) 238-246
• Margolis RH. 1985; Magnitude estimation of loudness. III: performance of selected hearing aid users. J Speech Hear Res 28 (03) 411-420
• Marozeau J, Florentine M. 2007; Loudness growth in individual listeners with hearing losses: a review. J Acoust Soc Am 122 (03) EL81-EL87
• Marriage JE, Moore BC, Stone MA, Baer T. 2005; Effects of three amplification strategies on speech perception by children with severe and profound hearing loss. Ear Hear 26 (01) 35-47
• Mencher GT, Clack TD, Rupp RR. 1973; Decruitment and the growth of loudness in the ears of brain-damaged adults. Cortex 9 (04) 335-345
• Moore B. 2003; Speech processing for the hearing-impaired: successes, failures, and implications for speech mechanisms. Speech Commun 41: 81-91
• Moore B. 2007. Cochlear Hearing Loss: Physiological, Psychological and Technical Issues. 2nd ed. Cambridge, England: John Wiley & Sons; 97-101
• Moore BC. 2001; Dead regions in the cochlea: diagnosis, perceptual consequences, and implications for the fitting of hearing AIDS. Trends Amplif 5 (01) 1-34
• Moore BC. 2004; a Dead regions in the cochlea: conceptual foundations, diagnosis, and clinical applications. Ear Hear 25 (02) 98-116
• Moore BC. 2004; b Testing the concept of softness imperception: loudness near threshold for hearing-impaired ears. J Acoust Soc Am 115 (06) 3103-3111
• Moore BC, Baer T, Ives DT, Marriage J, Salorio-Corbetto M. 2016; Effects of modified hearing aid fittings on loudness and tone quality for different acoustic scenes. Ear Hear 37 (04) 483-491
• Mueller H. 2003; Fitting test protocols are “more honored in the breach than the observance”. Hear J 56 (10) 19-26
• Mueller H. 2014 20Q: Real-ear probe-microphone measures – 30 years of progress?. Audiol Online. Retrived from: http://www.audiologyonline.com/articles/20q-probe-mic-measures-12410 . Accessed 11 October 2016
• Mueller H, Picou E. 2010; Survey examines popularity of real-ear probe-microphone measures. Hear J 63: 27-32
• Mueller H, Powers T. 2001; Consideration of auditory acclimatization in the prescriptive fitting of hearing aids. Semin Hear 22: 103-124
• Mueller HG, Bentler RA. 2005; Fitting hearing aids using clinical measures of loudness discomfort levels: an evidence-based review of effectiveness. J Am Acad Audiol 16 (07) 461-472
• Owens E, Schubert ED. 1977; Development of the California consonant test. J Speech Hear Res 20 (03) 463-474
• Padden C, Humphries T. 1990. Deaf in America: Voices from a Culture. Cambridge, MA: Harvard University Press; 91-104
• Padden C, Humphries T. 2005. Inside Deaf Culture. Cambridge, MA: Harvard University Press; 47-59 100–117; 170–178
• Palmer C. 2012 Siemens expert series: implementing a gain learning feature. Retrieved from: http://www.audiologyonline.com/articles/siemens-expert-series-implementing-gain-11244 . Accessed 11 February 2016
• Palmer C. 2013 Normal loudness perception: is it important, is it achievable?. Audiol Online. Retrieved from: http://www.audiologyonline.com/articles/siemens-expert-series-normal-loudness-11984 . Accessed 10 February 2016
• Palmer C, Lindley G. 2002. Overview and rationale for prescriptive formulas for linear and nonlinear hearing aids. In: Valente M. Strategies for Selecting and Verifying Hearing aid Fittings. Thieme Publishers; 1-22
• Prendergast S. 2005; Use of the California consonant test with children. J Educ Audiol 12: 67-75
• Priede VM, Coles RR. 1974; Interpretation of loudness recruitment tests--some new concepts and criteria. J Laryngol Otol 88 (07) 641-662
• Quar TK, Ching TY, Newall P, Sharma M. 2013; Evaluation of real-world preferences and performance of hearing aids fitted according to the NAL-NL1 and DSL v5 procedures in children with moderately severe to profound hearing loss. Int J Audiol 52 (05) 322-332
• Rasmussen AN, Olsen SO, Borgkvist BV, Nielsen LH. 1998; Long-term test-retest reliability of category loudness scaling in normal-hearing subjects using pure-tone stimuli. Scand Audiol 27 (03) 161-167
• Ricketts TA, Bentler RA. 1996; The effect of test signal type and bandwidth on the categorical scaling of loudness. J Acoust Soc Am 99 4 Pt 1 2281-2287
• Sanders J, Stoody T, Weber J, Mueller G. 2015; Manufacturers’ NAL-NL2 fittings fail real-ear verification. Hear Rev 21 (03) 24
• Schum D. 2009; Fitting severe and profound hearing losses with advanced technologies. Advance Audiologists 11 (02) 34-38 Retrieved from http://audiology.advanceweb.com/Editorial/Content/PrintFriendly.aspx?CC=195328
• Schwartz DM, Surr RK. 1979; Three experiments on the California consonant test. J Speech Hear Disord 44 (01) 61-72
• Scollie S, Seewald R, Cornelisse L, Moodie S, Bagatto M, Laurnagaray D, Beaulac S, Pumford J. 2005; The desired sensation level multistage input/output algorithm. Trends Amplif 9 (04) 159-197
• Scollie SD, Ching TY, Seewald RC, Dillon H, Britton L, Steinberg J, King K. 2010; Children’s speech perception and loudness ratings when fitted with hearing aids using the DSL v.4.1 and the NAL-NL1 prescriptions. Int J Audiol 49 (Suppl. 1) S26-S34
• Sherlock LP, Formby C. 2005; Estimates of loudness, loudness discomfort, and the auditory dynamic range: normative estimates, comparison of procedures, and test-retest reliability. J Am Acad Audiol 16 (02) 85-100
• Shi L-F, Doherty KA, Zwislockit JJ. 2007; Aided loudness growth and satisfaction with everyday loudness perception in compression hearing aid users. J Am Acad Audiol 18 (03) 206-219
• Silva I, Epstein M. 2012; Objective estimation of loudness growth in hearing-impaired listeners. J Acoust Soc Am 131 (01) 353-362
• Smeds K. 2004; Is normal or less than normal overall loudness preferred by first-time hearing aid users?. Ear Hear 25 (02) 159-172
• Smeds K, Keidser G, Zakis J, Dillon H, Leijon A, Grant F, Convery E, Brew C. 2006; Preferred overall loudness. II: listening through hearing aids in field and laboratory tests. Int J Audiol 45 (01) 12-25
• Sockalingam R, Lundh P, Schum D. 2011; Severe to profound hearing loss: what do we know and how do we manage it?. Hear Rev 18 (01) 30-33
• Souza PE, Jenstad LM, Folino R. 2005; Using multichannel wide-dynamic range compression in severely hearing-impaired listeners: effects on speech recognition and quality. Ear Hear 26 (02) 120-131
• Stach B. 1998. Clinical Audiology: An Introduction (Chapter 3: The Nature of Hearing Impairment. pp. 89–116) . San Diego, CA: Singular;
• Tharpe AM, Fino-Szumski MS, Bess FH. 2001; Survey of hearing aid fitting practices for children with multiple impairments. Am J Audiol 10 (01) 32-40
• Valente M, Abrams H, Benson D, Chisholm T, Citron D, Hampton D, Loavenbruck A, Ricketts T, Solodar H, Sweetow R. 2006; Guidelines for the audiologic management of adult hearing impairment. Audiol Today 18 (05) 1-44
• Vermeulen AM, Beijk CM, Brokx JP, van den Borne S, van den Broek P. 1995; Development of speech perception abilities of profoundly deaf children: a comparison between children with cochlear implants and those with conventional hearing aids. Ann Otol Rhinol Laryngol Suppl 166: 215-217
• Vickers DA, Moore BC, Baer T. 2001; Effects of low-pass filtering on the intelligibility of speech in quiet for people with and without dead regions at high frequencies. J Acoust Soc Am 110 (02) 1164-1175
• Villchur E, Killion MC. 2008; Measurement of individual loudness functions by trisection of loudness ranges. Ear Hear 29 (05) 693-703