380

J Am Acad Audiol 18:380–390 (2007)

*National Acoustic Laboratories, Chatswood, New South Wales, Australia; †Cooperative Research Centre for CochlearImplants and Hearing Aid Innovation, Melbourne, Australia

Ms. W. Pearce, National Acoustic Laboratories, 126 Greville St., Chatswood, NSW 2067 Australian Hearing; Phone +619412 6842; Fax: +61 9411 8273; E-mail: wendy.pearce@hearing.com.au

Cortical Auditory Evoked Potentials inthe Assessment of Auditory Neuropathy:Two Case Studies

Wendy Pearce*†Maryanne Golding*†Harvey Dillon*†

Abstract

Infants with auditory neuropathy and possible hearing impairment are beingidentified at very young ages through the implementation of hearing screeningprograms. The diagnosis is commonly based on evidence of normal cochlearfunction but abnormal brainstem function. This lack of normal brainstemfunction is highly problematic when prescribing amplification in young infantsbecause prescriptive formulae require the input of hearing thresholds that arenormally estimated from auditory brainstem responses to tonal stimuli. Withoutthis information, there is great uncertainty surrounding the final fitting. Corticalauditory evoked potentials may, however, still be evident and reliably recordedto speech stimuli presented at conversational levels. The case studies of twoinfants are presented that demonstrate how these higher orderelectrophysiological responses may be utilized in the audiological managementof some infants with auditory neuropathy.

Key Words: auditory neuropathy, cortical auditory evoked potentials, hearingaids

Abbreviations: ABR = auditory brainstem response; AN = auditory neuropathy;CAEPs = cortical auditory evoked potentials; CMs = cochlear microphonics;DPOAEs = distortion product otoacoustic emissions.; ECochG =electrocochleography; IHC = inner hair cell; NAL-NL1 = National AcousticLaboratories; OAEs = otoacoustic emissions; PBK = phonetically balancedkindergarten

Sumario

Palabras Clave:

Abreviaturas:

Background/Bratt et al

381

The term “auditory neuropathy” (AN)first appeared in audiological litera-ture in the mid-1990s (Sininger et al,

1995) although findings consistent with ANhad been reported for some years beforethis time (Worthington and Peters, 1980;Kraus et al, 1984; Widen et al, 1995). Oneof these early studies reported cases wherethe auditory brainstem response (ABR) wasabsent but behavioral thresholds had beenobtained in the absence of other neu-ropathology (Kraus et al, 1984), and theother study reported children with speechdelay who were found to have ABR resultsthat did not correlate with behavioralaudiometric data (Worthington and Peters,1980). In more recent times it has becomeclear that the term “AN” is too broad, orpossibly inappropriate in some cases, andso other terms such as “auditory dys-syn-chrony” (Berlin et al, 2002), “auditory de-synchrony” (Ray et al, in preparation) or“neural hearing loss” (Rapin and Gravel,2003) have also come into use.

The diagnosis of AN is commonly madewhen normal cochlear function (i.e., otoa-coustic emissions [OAEs] and/or a cochlear

microphonic [CM] are present) but abnor-mal brainstem function is evident (i.e.,absent, elevated, or grossly abnormal ABRresults) (Starr et al, 1996; Berlin et al,2003). Despite the disruption to the ABR, ithas been reported that behavioral thresh-olds measured in people with AN rangefrom normal (Kraus et al, 2000) to profoundand any degree of behavioral hearing lossin between (Rance et al, 1999). In addition,when speech perception results from openset speech tests are compared between peo-ple with AN and those with sensorineuralhearing loss measured behaviorally, of sim-ilar degrees, performance outcomes may beequal or results for those with AN may bemuch poorer (Rance et al, 2002).

While the term “neuropathy” refers topathology of peripheral nerve fibers, ANmay result from any one of a number of dis-orders or combination of them. It has beensuggested that the site of lesion may be atthe inner hair cells (IHC) and/or thesynapse with Type 1 auditory nerve fibers(Foerst et al, 2006). The disorder may alsoresult from a breakdown of the simultane-ous transmitter release from vesicles

attached to the synaptic ribbon of the IHCs,which results in impaired timing of afferentneuron firing (Fuchs et al, 2003; Khimich etal, 2005). AN may also arise if demyeliza-tion of Schwann cells around the auditorynerve fibers and/or alterations to nerveaxons impair the normal synchrony ofaxonal conduction velocity (Starr et al,1996; Rapin and Gravel, 2003) . This dis-ruption to peripheral function, which oftenleads to poor ABRs, does not necessarilyaffect cortical auditory evoked potentials(CAEPs) as these later responses are not asreliant on timing as the earlier evokedresponses (Hood, 1998; Rapin and Gravel,2003). Rance et al (2002) reported that in asample of 18 children diagnosed with AN,CAEPs were present in 50% of cases. Themere fact that some children with AN hadCAEPs and some did not lends increasedsupport to the hypothesis that AN describesa number of auditory dysfunctions andunderlying conditions and should not bethought of as a single disorder.

CAEPs have been recorded since the1960s although its popularity as a clinicaltool waned with the advent of ABR tech-niques. Davis (1966) reported that CAEPthresholds for tonal stimuli are within 10dB of behavioral thresholds in 90% of nor-mal-hearing and hearing-impaired adultsand children with a sensory loss. CAEPs toa variety of speech elements are also robustin normal-hearing infants, at least at con-versational levels, but using this techniqueto estimate threshold in infants is problem-atic as keeping them awake, but quietenough to elicit a threshold response, is dif-ficult (Cone-Wesson and Wunderlich, 2003).They have also been recorded in childrenand infants with hearing impairment, todemonstrate the detection of speech stimuliat the cortical level after hearing aid fitting(Rapin and Graziani, 1967; Gravel et al,1989), but this is by no means a routineclinical application.

Hearing aid fitting in infants that isbased on a prescriptive fitting formularequires hearing thresholds to be estimatedand entered into the formula. Tone-burstABR thresholds are typically used as theestimates, but when a diagnosis of AN hasbeen made, there is great uncertainty aboutthe appropriateness of using these values.More generally, there is uncertainty aboutthe application of a prescription rule that

has been derived based on the characteris-tics of people with sensorineural hearingloss to people with AN. Hearing aid fittingfor these children should therefore also bedependent on all available informationincluding behavioral test results and thefamily’s perspective on when the habilita-tion process should commence (King et al,2005).

Since the implementation of newbornhearing screening in the state of New SouthWales, there has been an increase in thenumber of infants diagnosed with AN andsubsequently referred to AustralianHearing for hearing habilitation. Thisreport presents the case studies of twoinfants who failed their newborn hearingscreening and were referred to experiencedpediatric audiologists for diagnostic audio-logical assessments at state hospitals andtest facilities. These assessments consistedof high-frequency probe-tone tympanome-try, tone burst ABR, and OAEs, and on thebasis of these results, the diagnosis of ANwas made. On referral to AustralianHearing, they were assessed with CAEPs toprovide additional guidance in the habilita-tion process.

METHOD

Procedure

The test stimuli were /m/ (duration 78msec), /g/ (duration 31 msec), and /t/ (dura-tion 78 msec), which were presented usingalternating onset polarity at typical conver-sational levels (i.e., 65 dB SPL or 75 dBSPL), and an interstimulus interval of 1125msec. These stimuli were generated fromnatural speech tokens consisting of an ini-tial consonant followed by the vowel /ae/,which was extracted from a recording ofrunning speech that was spoken by anaverage male Australian. The frequencyresponse of the final test stimuli is shownin Figure 1. They included very little of thevowel transition and were recorded withdigitization rates of 40 kHz. They weregated off at a zero crossing to minimizeaudible clicks, and no further modificationsof the onset or offset characteristics weremade. These consonants were chosenbecause they had a spectral emphasis inthe low, mid-, and high-frequency regions,

Journal of the American Academy of Audiology/Volume 18, Number 5, 2007

382

respectively, and thus had the potential togive diagnostic information about the per-ception of speech sounds in different fre-quency regions. Prior to testing, the stimulioutputs at the sound-field test positionwere measured as 65 dB SPL and 75 dBSPL (impulse time constant) using a micro-phone suspended from the ceiling and con-nected to a measuring amplifier in theobservation room. The microphone wasthen retracted to a point above head heightfor continuous monitoring of the signal.

Brain electrical activity was recordedusing the NeuroscanTM system with elec-trodes positioned at Cz, C3, and C4 refer-enced to one mastoid with forehead asground. During cortical testing, infantswere awake and seated on their mother’slap or in a baby chair, distracted by anoth-er adult if required. Stimuli were deliveredvia a loudspeaker positioned at 45 degreesazimuth on either side of the subject. Ifunaided, stimuli were presented throughthe speaker to the left side as the defaultsetting. If aided, the speaker nearest thetest ear was used while the opposite earwas occluded by the child’s own earmoldand hearing aid in the off position.Individual sweeps of the electroencephalicactivity were amplified and analog filteredonline at 0.1–100 Hz using a 24 dB/octave

slope and subsequently filtered off line at1–30 Hz. The recording window consisted ofa 100 msec pre-stimulus baseline and a fur-ther 600 msec. Artifact reject was set at±150 µV.

In keeping with our standard protocols,each stimulus was presented in blocks until100 artifact-free EEG samples wereacquired and, where possible, each block ofstimuli was presented on two occasionswith a randomized stimulus order. If theblock was not repeated because the infantgrew tired of testing, responses in the sin-gle stimulus block were separately aver-aged for the odd and even stimulus presen-tations. Response detection was based onthe two replicated waveforms being over-laid and inspected for repeatability by anexaminer who was experienced in identify-ing infant CAEPs. For the purposes of thesecase studies, no attempt was made to markpeak latency or amplitude.

CASE STUDIES

Infant 1

Infant 1 was born at 28 weeks gestationwith low birth weight, a poor apgar score,jaundice, and respiratory distress. He wasseen for diagnostic audiology at 12 weeks of

Title/Author

383

Figure 1. Spectral analysis of the three stimuli /g/, /t/, and /m/ showing the primary energy of /m/ below 400 Hz,/g/ between 1200 and 4000 Hz, and /t/ above 3000 Hz.

age (i.e., 40 weeks gestational age) by expe-rienced pediatric audiologists at a largeteaching hospital. The clinicians reportedthat there were no responses to tone-burstABR at 500 Hz and 2000 Hz at the limits ofthe equipment (i.e., 85 dB nHL for 500 Hzand 100 dB nHL for 2000 Hz), but distor-tion product OAEs (DPOAEs) were presentacross all frequencies bilaterally. Althoughthe infant was referred to AustralianHearing for habilitation on the basis ofthese results, the family reported that theinfant was responsive to sounds at home

such as familiar voice and the stereo, andso a repeat ABR was performed approxi-mately six weeks later at the hospital. Atthis second visit, click stimuli were pre-sented with a possible threshold of 95 dBnHL achieved in both ears.

Unaided CAEP testing was performedone week later (i.e., at seven weeks cor-rected age). Repeatable responses for allthree speech stimuli (i.e., /m/, /g/, /t/) wereobserved with presentation levels of 65 dBSPL as shown in Figure 2. This resultshows that responses to CAEPs were

Journal of the American Academy of Audiology/Volume 18, Number 5, 2007

384

Figure 2. Repeatable CAEP responses for infant 1 are shown in response to the three speech stimuli:(A) stimulus /m/, (B) stimulus /g/, (C) stimulus /t/.

obtained at significantly lower stimuluslevels than for ABR. A severe hearing losscould be excluded as behavioral responsesand cortical responses were seen at mod-erate levels but no attempt was made toestablish threshold using CAEP testing. Amild to moderate loss could not, however,be excluded, and therefore, hearing aidswere fitted using Australian Hearing rec-ommendations (King et al, 2005). In brief,these recommendations advise that allavailable measures of auditory function,including electrophysiological and behav-ioral test results, be considered along withparental desire to proceed with the fitting.On the basis of the CAEP results and thisinfant’s parental observations, and toavoid potential overprescription of thehearing aid gain, the audiogram was esti-mated at 30 dB at 500 Hz, 40 dB at 1000Hz, 50 dB at 2000 Hz, and 50 dB at 4000Hz. This audiogram was entered to NAL-NL1 (National Acoustic Laboratories—Nonlinear 1) prescription formula, and hewas fitted with digital behind-the-earhearing aids using wide dynamic rangecompression and monitored regularly byhis clinician. With this degree of amplifica-tion, there was no possibility of damage tothe cochlea even if the auditory thresholdsproved to be normal.

Given the ongoing uncertaintiesregarding the infant’s final thresholds,electrocochleography (ECochG) was car-ried out five months later by the infant’sphysician. There were no responses at 110dB nHL to tonal stimuli, but CAEP test-ing two weeks later again showed repeat-able responses for all three speech stimuli(i.e., /m/, /g/, /t/) with presentation levelsof 65 dB SPL

Regular adjustments were made to thefitting based on parental report, the out-comes of behavioral tests, and anotherunaided CAEP test, which continued todemonstrate repeatable responses at 65dB SPL. At the age of two-and-a-halfyears, behavioral responses to visual rein-forcement orientation audiometry wereobtained at 30 dB SPL at 500 Hz, 25 dBSPL at 1000 Hz, 30 dB SPL at 2000 Hz,and 30 dB SPL at 4000 Hz. The child wasno longer using hearing aids and wasenrolled in an early learning program dueto delay in his speech development.

Infant 2

This infant was born at full term with theonly risk factor being a family history of twopaternal relatives who used cochlearimplants. Their etiology is unknown. He wasdiagnosed with AN at ten weeks of age byexperienced pediatric audiologists at a special-ist diagnostic facility. Initial ABR tests, usingclick stimuli, showed no response at the max-imum limit of 100 dB nHL for the left ear anda repeatable wave V response at 100 dB nHLfor the right ear as shown in Figure 3. ClearCMs and robust DPOAEs were, however, seenfor both ears. In keeping with this clinic’s pol-icy for infants with AN, a repeat ABR wasordered with consultations for habilitation atAustralian Hearing and medical revieworganized in the interim. He was fitted withdigital hearing aids at four months of ageusing Australian Hearing protocols (King et

Title/Author

385

Figure 3. ABR s to click stimuli are shown for infant 2:(A) right ear, (B) left ear.

al, 2005). A conservative audiogram that pre-dicted a moderate loss was entered into theNAL-NL1 prescription in the first instance asthere were no clear behavioral responses toauditory stimuli and limited electrophysiolog-ical information.

A click and tone burst ABR was performedthree months after the first test and showedno response to click stimuli in the left ear atmaximum output, although cochlear micro-phonic activity was still reported to be evident.Testing for the right ear was limited to two fre-

quencies as the infant did not sleep well. Noresponse to 4000 Hz at 105 dB nHL or 2000Hz at 90 dB nHL was observed, which wasconsistent with the results from the first test.

Aided CAEP testing was performed twoweeks after the second ABR test. Parents feltthat their baby was a little more responsivewith his hearing aids on than without them,but there were very few clear examples of audi-tory behavior. The aided CAEP test results for65 dB SPL presentation levels are shown inFigures 4 and 5.

Journal of the American Academy of Audiology/Volume 18, Number 5, 2007

386

Figure 4. Nonrepeatable CAEP responses for infant 2 (left ear aided) are shown in response to the threespeech stimuli: (A) stimulus /m/, (B) stimulus /g/, (C) stimulus /t/.

There were no repeatable corticalresponses to any of the three speech stimulipresented at 65 dB SPL to the left or theright ear when aided. Limited testing wasperformed at 75 dB SPL as the infantbecame restless, with no response observedin either ear to /t/ at 75 dB SPL. The infant’shearing aids were subsequently adjusted toprovide greater gain by re-estimating thedegree of hearing loss and recalculating theNAL-NL1 prescription. Aided CAEP testingwas performed a second time after the

adjustment, but no response to stimulationat 75 dB SPL could be observed on this sec-ond occasion, and therefore, the hearingaids were further adjusted to provide theirmaximum gain. CAEP testing was repeatedwithin a few days of this final adjustment,but still no response could be observed.

The child underwent a series of tests andfor cochlear implantation. He exhibited asevere hearing loss using ECochG, andthere was evidence of auditory nerve func-tion using electrically evoked ABR testing.

Title/Author

387

Figure 5. Nonrepeatable CAEP responses for infant 2 (right ear aided) are shown in response tothe three speech stimuli: (A) stimulus /m/, (B) stimulus /g/, (C) stimulus /t/.

There were no inner ear abnormalitiesfound on computerized tomography andmagnetic resonance imaging studies. Afterextensive family counseling to ensure thatthe results were understood and that a pos-itive commitment to the process existed, hereceived a cochlear implant at one year andfour months of age.

DISCUSSION

Hearing aid prescription for younginfants is often based on thresholds

that have been estimated from frequency-specific ABR recordings (Dillon, 2001). Incases where AN is present, this approach ishighly problematic as the ABR response isoften poorly formed or absent and poten-tially misleading with behavioral hearingthresholds much better than those estimat-ed by the ABR results. The possibility ofoverprescribing the gain requirementstherefore exists if decisions are based onABR thresholds alone. Although no attemptwas made to find CAEP threshold in eithercase presented in this report, the detectionor lack of CAEP responses to speech stimulipresented at conversational levels assistedin clinical decision making for these twochildren.

In the case of our first infant, it was clearthat the three speech stimuli could bedetected at the level of the cortex at conver-sational level without the assistance ofamplification. Although it was not possibleto infer that hearing thresholds were nor-mal in one or both ears, these results cer-tainly suggested that the degree of hearingloss was much less than that suggested bythe ABR results. It has been reported(Berlin et al, 2002) that premature infants,or those suffering with hyperbilirubinemia,may have an initial presentation normallyassociated with AN, but they are more like-ly to recover with time. Our first infant wasvery premature when born at 28 weeks andhad poor ABR test outcomes on two occa-sions over the first 18 weeks of life. It isquite possible that the ABR may haveimproved over time, and it would have beenvaluable to observe this improvement, butrelying on ABR results alone in these earlymonths would have prolonged the period ofanxiety for parents and clinicians andwould have resulted in a fitting inconsis-tent with the child’s ability to detect

sounds. This child was fitted with hearingaids set for a mild-to-moderate hearing lossas suggested by the CAEP results, and theensuing clinical decisions and modificationsto the fit were based on monitoring thechild’s auditory behavior and parentalreport, as well as CAEP testing. Sometimelater, behavioral testing was able to confirmthat hearing thresholds, in the sound field,were normal. This outcome did not, howev-er, guarantee that the child had goodspeech perception skills. Kurtzberg (1989)reported that infants with normal CAEPresponses were more likely to show normalreceptive language at the age of one year.This child’s parents reported informallythat he had good receptive skills at two-and-one-half years of age (e.g., he could fol-low instructions well), but speech produc-tion was very poor.

Rance et al (2002) found that in childrenwith AN, the presence or absence of CAEPsat presentation levels of 20 to 40 dB SL waspositively correlated with aided phonemicscores on the phonemically balancedkindergarten (PBK) word tests. Childrenwith CAEPs had an average aided PBKscore of 60%, and those without CAEPsshowed an average score of 6%. Theauthors concluded that the recording ofCAEPs might therefore be a means of pre-dicting speech perception skills. Lee et al(2001), however, described two childrenwho had OAEs and attended a school forhearing-impaired children. They had poorABRs but clear CAEPs. Both these chil-dren, who had moderate degrees of hearingloss behaviorally, had rejected their hearingaids and had poor speech-discriminationscores. Hood (1999) also reported the caseof an adult diagnosed with Charcot-Marie-Tooth syndrome who similarly had clearCAEPs, robust OAEs, and poor ABRresponses. This adult, who had a moderate-to-severe hearing loss behaviorally, did notfind hearing aids helpful and was reliant onlip reading and other visual cues for com-munication. It appears, then, that in casesof AN, the presence of a clear corticalresponse at suprathreshold presentationlevels may not always be predictive of goodspeech perception skills. Children with ANshould, therefore, be regularly monitoredfor speech-language delay and interventionorganized as appropriate.

Our second infant had absent ABRs and

Journal of the American Academy of Audiology/Volume 18, Number 5, 2007

388

absent aided CAEPs at conversational lev-els. The infant was therefore fitted usinginfant protocols for severe-to-profoundhearing loss and regular monitoring of thechild’s performance both behaviorally andwith CAEPs was undertaken. No aidedCAEP could be recorded at conversationallevels, even after several incremental gainadjustments were made, and improvementsto auditory behavior could not be observed.As a result he was evaluated for cochlearimplantation at one year of age.

There are of course many children withAN who have some ABR response albeitdistorted and potentially well above hear-ing threshold. In such cases, combinedinformation from CAEP and ABR testingmay be valuable in clinical decision mak-ing. It is also acknowledged that establish-ing CAEP thresholds may provide a usefuladditional perspective on the child’s hear-ing, but it remains unclear whether audito-ry thresholds can be estimated reliablyusing this technique in all but the calmestof infants (Cone-Wesson and Wunderlich,2003). Even without this extra information,the mere presence/absence of a CAEPresponse to speech stimuli at conversation-al level can provide useful clinical informa-tion. It is our contention that, for infantswith AN, hearing aids might be fitted basedon an assumed mild-to-moderate hearingthreshold when unaided CAEPs are evidentin response to speech stimuli presented atconversational levels. It is important, how-ever, to arrange regular formal monitoringof performance using CAEPs, parentalquestionnaire, and behavioral testing. IfABRs and aided CAEPs are absent, howev-er, a more aggressive approach to interven-tion may be warranted.

Acknowledgments. We thank colleagues at the JimPatrick Audiology Centre at the Royal Institute forDeaf and Blind Children, and the New Children’sHospital Westmead for access to the diagnostic testresults for these two infants.

Title/Author

389

REFERENCES

Berlin C, Li L, Hood LJ, Morlet T, Rose K, BrashearsM. (2002) Auditory neuropathy/dys-synchrony: afterthe diagnosis, then what? Semin Hear 23(3):29–214.

Berlin C, Morlet T, Hood L. (2003) Auditory neu-ropathy/dys-synchrony. Its diagnosis and management.Pediatr Clin North Am 50(2):331–340.

Cone-Wesson B, Wunderlich J. (2003) Auditory evokedpotentials from the cortex: audiology applications.Curr Opin Otolaryngol Head Neck Surg 11:372–377.

Davis H. (1966) Validation of evoked-response audiom-etry (ERA) in deaf children. Int Audiol 5(2):77–81.

Dillon H. (2001) Hearing Aids. New York: ThiemePublishers.

Foerst A, Beutner D, Lang-Roth R, Huttenbrink K-B,Von Wedel H, Walger M. (2006) Prevalence of audi-tory neuropathy/synaptopathy in a population ofchildren with profound hearing loss. Int J PediatrOtorhinolaryngol 70:1415–1422.

Fuchs P, Glowatzki E, Moser T. (2003) The afferentsynapse of cochlear hair cells. Curr Opin Neurobiol13:452–458.

Gravel JS, Kurtsberg D, Stapells DR, Vaughan HG,Wallace IF. (1989) Case studies. Semin Hear10(3):272–287.

Hood L. (1998) Auditory neuropathy: what is it andwhat can we do about it? Hear J 51:10–18.

Hood L. (1999) A review of objective methods of eval-uating auditory neural pathways. Laryngoscope109:1745–1748.

Khimich D, Nouvian R, Pujol R, Tom Dieck S, EgnerA. (2005) Hair cell synaptic ribbons are essential forsynchronous auditory signaling. Nature434(7035):889–894.

King A, Purdy S, Dillon H, Sharma M, Pearce W. (2005)Australian Hearing protocols for the audiological man-agement of infants who have auditory neuropathy.Aust N Z J Audiol 27(1):69–77.

Kraus N, Azdamar A, Stein L, Reed N. (1984) Absentauditory brainstem response: peripheral hearing lossor brainstem dysfunction? Laryngoscope 94:400–406.

Kraus N, Bradlow A, Cheatham M, Cunningham J,King C, Koch D, Nichol T, Mc Gee T, Stein L, WrightB. (2000). Consequences of neural asynchrony: a caseof auditory neuropathy. J Assoc Res Otolaryngol1:33–45.

Kurtzberg D. (1989) Cortical event-related potentialassessment of auditory system function. Semin Hear10(3):252–262.

Lee JS, McPherson B, Yuen KC, Wong LL. (2001)Screening for auditory neuropathy in a school of hear-ing impaired children. Int J Pediatr Otorhinolaryngol61:39–46.

Rance G, Beer D, Cone-Wesson B, Shepherd R, DowellR, King A, Rickards F, Clark G. (1999) Clinical find-ings for a group of infants and young children withauditory neuropathy. Ear Hear 20:238–252.

Rance G, Cone-Wesson B, Wunderlich J, Dowell R.(2002) Speech perception and cortical event relatedpotentials in children with auditory neuropathy. EarHear 23:239–253.

Rapin I, Gravel J. (2003) “Auditory neuropathy”: phys-iologic and pathologic evidence calls for more diagnosticspecificity. Int J Pediatr Otorhinolaryngol 67:707–728.

Ray J, Gibson W, Sanli H, Haddon A. (In preparation)Brainstem auditory neuropathy, hair cell desynchronyand cochlear implantation.

Sininger Y, Hood L, Starr A, Berlin C, Picton T. (1995)Hearing loss due to auditory neuropathy. Audiol Today7:10–13.

Starr A, Picton T, Sininger Y, Hood L, Berlin C. (1996)Auditory neuropathy. Brain 119:741–753.

Widen J, Ferraro J, Trouba S. (1995) Progressiveneural hearing impairment: case report. J Am AcadAudiol 6:217–224.

Worthington D, Peters J. (1980) Quantifiable hearingand no ABR: paradox or error? Ear Hear 1:281–285.

Journal of the American Academy of Audiology/Volume 18, Number 5, 2007

390