Can Routine Office-Based Audiometry Predict Cochlear Implant Evaluation Results?

Transcription

1 The Laryngoscope VC 2016 The American Laryngological, Rhinological and Otological Society, Inc. TRIOLOGICAL SOCIETY CANDIDATE THESIS Can Routine Office-Based Audiometry Predict Cochlear Implant Evaluation Results? Samuel P. Gubbels, MD, FACS; Brian C. Gartrell, MD; Jennifer L. Ploch, MA, CCC-A; Kevin D. Hanson, BS Objectives/Hypothesis: Determining cochlear implant candidacy requires a specific sentence-level testing paradigm in best-aided conditions. Our objective was to determine if findings on routine audiometry could predict the results of a formal cochlear implant candidacy evaluation. We hypothesize that findings on routine audiometry will accurately predict cochlear implant evaluation results in the majority of candidates. Study Design: Retrospective, observational, diagnostic study. Methods: The charts of all adult patients who were evaluated for implant candidacy at a tertiary care center from June 2008 through June 2013 were included. Routine, unaided audiologic measures (pure-tone hearing thresholds and recorded monosyllabic word recognition testing) were then correlated with best-aided sentence-level discrimination testing (using either the Hearing in Noise Test or AzBio sentences test). Results: The degree of hearing loss at 250 to 4,000 Hz and monosyllabic word recognition scores significantly correlated with sentence-level word discrimination test results. Extrapolating from this association, we found that 86% of patients with monosyllabic word recognition scores at or below 32% (or 44% for patients with private insurance) would meet candidacy requirements for cochlear implantation. Conclusions: Routine audiometric findings can be used to identify patients who are likely to meet cochlear implant candidacy upon formal testing. For example, patients with pure-tone thresholds (250, 500, 1,000 Hz) of 75 db and/or a monosyllabic word recognition test score of 40% have a high likelihood of meeting candidacy criteria. Utilization of these predictive patterns during routine audiometric evaluation may assist hearing health professionals in deciding when to refer patients for a formal cochlear implant evaluation. Key Words: Cochlear implant, adult, hearing loss, criteria, evaluation, candidacy, audiometry. Level of Evidence: 4 Laryngoscope, 127: , 2017 INTRODUCTION The effectiveness of cochlear implantation has been well documented, including improvements in quality of life, 1 psychological well-being, 2 self-esteem, 3 and reduction in stress and tinnitus impairment. 4,5 Furthermore, the improvements in speech understanding and communication provided by a cochlear implant (CI) have been shown to ameliorate social isolation 3,6 and improve productivity at work. 5,7 In addition, unilateral cochlear implantation has been associated with an economic benefit, with cost-utility ratios ranging from $14,000 to $16,000 per quality-adjusted life-year, thus making it a highly cost-effective procedure Although the benefits of cochlear implantation have been well described, it is clear that many individuals with hearing loss who would qualify for and receive benefit from this technology do not receive an implant. 12,13 For example, some sources suggest that <6% of adults Additional Supporting Information may be found in the online version of this article. From the Department of Otolaryngology (S.P.G.), University of Colorado, School of Medicine, Aurora, Colorado; Department of Surgery, Division of Otolaryngology (S.P.G., B.C.G.), University of Wisconsin Madison, School of Medicine and Public Health, Madison, Wisconsin; Department of Family Medicine, University of Nebraska Medical Center (B.C.G.), Omaha, Nebraska; Department of Surgery, Division of Audiology (J.L.P.), University of Wisconsin Madison, School of Medicine and Public Health, Madison, Wisconsin; and the University of Wisconsin Madison, School of Medicine and Public Health (K.D.H.), Madison, Wisconsin, U.S.A. Editor s Note: This Manuscript was accepted for publication March 28, Portions of the data from this study were presented at the American Academy of Otolaryngology Head and Neck Surgery Annual Meeting, Orlando, Florida, U.S.A., September 19 23, This report was a submitted to and accepted by the Triological Society as a Triological Thesis (for author S.P.G.) (thesis number ). S.P.G. receives funding from the National Institutes of Health, National Institute on Deafness and Other Communicative Disorders (1 R01 DC ) for research unrelated to this report. S.P.G. received funding (KL2) from the Clinical and Translational Science Award program by the National Center for Advancing Translational Sciences, grant TL1 UL1TR K.D.H. received funding from the University of Wisconsin School of Medicine and Public Health, Shapiro Summer Research Program. This project was supported by research funds from the University of Wisconsin Madison School of Medicine and Public Health, Department of Surgery, Division of Otolaryngology. The authors have no other funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Samuel Gubbels, MD, East 17th Avenue, MS-B205, Aurora, CO samuel.gubbels@ucdenver.edu DOI: /lary

2 in the United States who could potentially receive benefit from a CI have been implanted. 14 Although there are many reasons for the underutilization of this technology, one contributory factor is that there are no established criteria on routine office-based audiometry to assist hearing health providers (audiologists, hearing instrument specialists, otolaryngologists) in their decisions on whether to refer patients for a formal CI evaluation. Consequently, it can be challenging for hearing health professionals to make appropriate referrals for CI evaluation based upon the testing data that they have available. Ideally there would be a method to accurately identify potential candidates using standard audiometric testing. Having such a method would be of value to cochlear implantation centers and patients alike, as it would help to select those with a high likelihood of meeting candidacy requirements while avoiding the resource expenditures, travel, and patient frustration associated with a negative candidacy evaluation. We were unable to find any relevant studies addressing this topic in the existing literature. Our hypothesis is that elements of a routine audiogram will correlate strongly with the findings on formal CI candidacy evaluation. If strong correlation between routine audiometry and the results of formal CI evaluation are present, then extrapolation of current CI candidacy requirements can be performed. Doing so would then allow for identification of thresholds of hearing loss on routine audiometry that can be used to guide the CI referral process in the most costeffective and efficient manner. Data from this study could serve as a useful tool to help guide referrals for cochlear implantation. MATERIALS AND METHODS Subjects All adults who were evaluated for cochlear implantation at a tertiary care medical center cochlear implantation program for a 5-year period starting May 2008 were included. Data were collected retrospectively. Exclusion criteria included incomplete audiometric datasets, elective implantation without insurance approval, patients evaluated for revision or reimplantation procedures, and patients whose initial evaluations fell outside of the approved study window. The University of Wisconsin Institutional Review Board approved this study. Audiological Performance Assessment Results from the subject s most recent routine hearing exam prior to CI evaluation were reviewed. Pure-tone hearing thresholds at 250, 500, 1,000, 2,000, 4,000, 6,000 and 8,000 Hz were recorded. When a conductive component was present, bone-conduction thresholds were recorded. The patient s pattern of hearing loss was characterized as downsloping, flat, or other. A downsloping hearing loss was defined as a predominantly high-frequency hearing loss wherein the pure-tone threshold values worsened by 30 db HL from low to high frequency. A flat pattern of hearing loss was defined as having similar reduction in pure-tone threshold values across all the tested frequencies, specifically having all the values within 30 db HL of each other. In addition to evaluation of pure-tone thresholds, monosyllabic word recognition test (MWRT) results were recorded, and the use of recorded or live voice was noted. Results from the patient s formal CI evaluation were then evaluated by recording the best-aided sentence-level word discrimination test (SWDT) results. The type of sentence-level test materials used (Hearing in Noise Test [HINT], AzBio) was recorded along with whether the test was conducted in quiet or in noise. All AzBio testing in noise was performed with 15 signal-to-noise (SNR) ratio. Recorded voice was used for all sentence-level testing. Statistical Analysis Analyses of correlation. To investigate the correlative value of routine audiometric measurements with the CI evaluation results, statistical analysis was performed using SPSS software (SPSS, Inc., Chicago, IL). Initial statistical analyses included performing Pearson correlations (Pc) that included the best ear pure-tone hearing threshold values at all frequencies, the MWRT results, and the SWDT results. Plots were then made correlating the best-ear pure-tone threshold and MWRT results with the SWDT scores. The data were then separated according to the pattern of hearing loss (defined above), and the same statistical analyses were performed for each subgroup (downsloping, flat, other). The significance level for the correlations performed was set at a <.05. Extrapolation of threshold values. We sought to determine the pure-tone and MWRT threshold values on routine audiometry at which individuals in our study group would be deemed candidates for a CI. Using the previously generated best-ear pure-tone threshold versus sentence-level word discrimination test result plots we extrapolated pure-tone CI candidacy thresholds. At each frequency, a threshold of 40.0% correct (Medicare candidacy criteria) on SWDT was set to establish a conservative candidacy threshold. Using the best-fit line, 40.0% correct on SWDT correlated to a certain frequencyspecific pure-tone threshold value (db HL), a value that could be set as a candidacy benchmark. The percentage of patients with hearing loss at or worse than the benchmark value at each frequency who met CI candidacy criteria was then determined. A similar analysis was performed on the best-ear MWRT versus SWDT scatter plots to determine a monosyllabic word recognition threshold for CI candidacy. RESULTS Subjects and Audiologic Testing From 2008 to 2013, 147 subjects were identified who underwent formal CI evaluations. Of those 147, eight were excluded for various reasons: one paid out of pocket for the procedure, one died prior to having adequate follow-up time, one later had the contralateral ear implanted, one had a prior labyrinthectomy in the implanted ear, and four had incomplete medical records. Of the 139 subjects included in our study, 70 were female (50.4%) and the average age was 63.1 years, ranging from 20 to 92 years old. In describing the subject s pattern of hearing loss, 75 were downsloping, 56 were flat, and eight were considered other (i.e., not flat or downsloping). Ultimately, 102 subjects qualified for implantation. In regard to which sentence material was used, 80 subjects were evaluated with HINT sentences, 52 with AzBio sentences, and seven with both. All of the MWRTs and SWDTs were completed using voice recordings. Analysis of Correlation Grouped pure-tone testing results at individual frequencies were evaluated for correlation with SWDT scores by generating a Pearson correlation coefficient 217

3 TABLE I. Individual Hearing Threshold Frequencies and Monosyllabic Word Recognition Testing Results Correlated to the Sentence Level Word Discrimination Testing Results. 250 Hz 500 Hz 1,000 Hz 2,000 Hz 4,000 Hz 6,000 Hz 8,000 Hz MWRT HINT (quiet) Pc * * * * * * Sig No AzBio (quiet) Pc 2.542* * * * * * Sig No AzBio (noise) Pc 2.599* * * * * Sig No *Correlations were considered significant at P <.05. HINT 5 Hearing in Noise Test; Pc 5 Pearson correlation; Sig. 5 significance; MWRT 5 mono-syllabic word recognition value. In this analysis, Pearson correlation coefficients between 0 and were not significant. Those between and 21 were significant, with values closer to 21 indicating higher levels of correlation. Our results demonstrate that pure-tone thresholds in the better ear at 250, 500, 1,000, and 2,000,Hz had significant correlation (P <.05) with sentence-level word discrimination testing results using both HINT and AzBio sentences (with or without noise) (Table I). Thresholds at 4,000 Hz had significant correlation (P <.05) with sentence-level testing results using HINT and AzBio sentences without noise. Pure-tone threshold values at 6,000 and 8,000 had poor or nonsignificant correlation with SWDT regardless of the material used. Similarly, we found a strong correlation of MWRT with SWDT in all test conditions. The strongest correlation of MWRT was with HINT sentence (Pc ), followed by AzBio in quiet (Pc ) then AzBio in noise (Pc ). Extrapolation of Threshold Values Using the current Medicare CI candidacy threshold of 40.0%, we established threshold values for relevant individual pure tones (frequencies with a significant Pc from above) (Fig. 1) (see also Supporting Figures 1 4 in the online version of this article). Our data show that by using this strategy, 83.3% of patients with pure tones at or below 54.7 db at 250 Hz would meet CI candidacy upon formal CI candidacy evaluation using AzBio Fig. 1. Individual pure-tone hearing threshold frequencies were correlated to sentence-level word discrimination testing results using AzBio sentences without background noise. At the 250 Hz frequency, 83.3% of patients who obtained 54.7 db or worse at 250 Hz during pure-tone threshold testing qualified for cochlear implantation using AzBio sentences without background noise. Fig. 2. A correlation between the monosyllabic word recognition testing results and the sentence-level word discrimination testing results, using AzBio sentences without background noise. Using Medicare criteria as a threshold, 81.5% of patients who scored 35.6% or worse on monosyllabic word recognition testing qualified for cochlear implantation. MWRT 5 monosyllabic word discrimination 218

4 Fig. 3. A correlation between the monosyllabic word recognition testing results and the sentence-level word discrimination testing results, using AzBio sentences with background noise. Using Medicare criteria as a threshold, 93.3% of patients who scored 39.2% or worse on monosyllabic word recognition testing qualified for cochlear implantation. MWRT 5 monosyllabic word discrimination Fig. 5. A correlation between the MWRT results and the SWDT results, regardless of the sentence material used. Using private insurance criteria as a threshold, 88.1% of patients who scored 44.2% or worse on the MWRT qualified. MWRT 5 monosyllabic word recognition test; SWDT 5 sentence-level word discrimination sentences (Fig. 1) in quiet. Data from this analysis for the pure-tone values of 500, 1,000, 2,000, and 4,000 are presented in Supporting Figures 1 through 4 in the online version of this article. Applying a similar analysis to the results of the MWRT versus SWDT, we found that 87.0% of patients who scored 30.2% or less correct in the MWRT qualified for a CI according to Medicare criteria using the HINT sentences (see Supporting Figure 5 in the online version of this article). Similarly, 81.5% of patients who scored 35.6% or less correct in the MWRT qualified using AzBio sentences in quiet (Fig. 2), and 93.3% of hearing loss patients who scored 39.2% or less correct in the MWRT qualified using AzBio sentences with background noise (Fig. 3). When the results of the SWDT were grouped together, regardless of the testing material used, we found that 85.5% of patients with MWRT scores 32.8% would qualify for a CI using Medicare candidacy criteria (40% or worse on sentence testing) (Fig. 4). Likewise, we found that 88.1% of patients with MWRT scores 44.2% would qualify when using private insurance candidacy criteria (60% or worse on sentence testing) (Fig. 5). Fig. 4. A correlation between the MWRT results and the SWDT results, regardless of the sentence material used. Using Medicare criteria as a threshold, 85.5% of patients who scored 32.8% or worse on the MWRT qualified for cochlear implantation. MWRT 5 monosyllabic word recognition test; SWDT 5 sentence-level word discrimination Subgroup Analysis: Pattern of Hearing Loss In subjects who demonstrated a downsloping pattern of hearing loss, 71.4% of patients who scored 30.0% or less correct in the MWRT qualified for a CI when using HINT sentences (see Supporting Figure 6 in the online version of this article). Furthermore, when using AzBio sentences in quiet, 68.8% of patients who scored 35.0% or less correct in the MWRT qualified for a CI (Fig. 6). When using AzBio sentences with background noise, 90.0% of patients who scored 39.0% or less on MWRT qualified for a CI (see Supporting Figure 7 in the online version of this article). For patients with a flat pattern of hearing loss, 96.7% of patients who scored 33.0% or less correct in the MWRT qualified for a CI when using HINT sentences (see Supporting Figure 8 in the online version of this article). When using AzBio sentences in quiet, 90.0% of patients who scored 40.6% or less correct in MWRT qualified for a CI (Fig. 7). Given the low number of patients with flat-patterned hearing losses tested using AzBio in noise, we were unable to determine with confidence a similar MWRT threshold value. 219

5 TABLE II. Summary of the Reference Threshold Values to Guide Cochlear Implant Evaluation Referral. Audiometric Measure Threshold Value Patients Meeting CI Candidacy (%) 250 Hz 54.7 db 83.3 (Medicare criteria) 500 Hz 62.5 db 79.3 (Medicare criteria) 1 KHz 75.7 db 80.0 (Medicare criteria) 2 KHz 88.6 db 81.5 (Medicare criteria) 4 KHz 92.6 db 79.3 (Medicare criteria) WRS (MWRT) 32.8% 85.5 (Medicare criteria) WRS (MWRT) 44.2% 88.1 (private insurance criteria) Fig. 6. Correlating performance on the MWRT to results in the SWDT for downsloping patterns of hearing loss. Using Medicare criteria as a threshold, 68.8% of patients who scored 35.0% or worse in the MWDT qualified using AzBio sentences without background noise. MWRT 5 monosyllabic word recognition test; MWDT 5 monosyllabic word discrimination test; SWDT 5 sentence-level word discrimination Fig. 7. Correlating performance on the MWRT to results in the SWDT for flat patterns of hearing loss. Using Medicare criteria as a threshold, 90.0% of patients who scored 40.6% or worse in MWDT qualified using AzBio sentences without background noise. MWRT 5 monosyllabic word recognition test; MWDT 5 monosyllabic word discrimination test; SWDT 5 sentence-level word discrimination The audiometric measures (left column) and corresponding threshold values (middle column) that significantly correlated with meeting candidacy upon cochlear implant evaluation are shown. The right column shows the numbers of patients in the study cohort who met cochlear implant candidacy requirements when at or below the threshold values listed based upon Medicare candidacy criteria (40% or worse on sentence level discrimination tests [upper six rows]) or using private insurance criteria (60% or worse on sentence level tests [lower row]). For pure-tone measures, the threshold values were generated through correlation with sentence level discrimination testing using AzBio in quiet. For the word recognition scores (MWRT) the threshold values were generated through correlation with sentence-level discrimination testing regardless of the type of material used (AzBio in quiet, AzBio in noise, or HINT in quiet). For example, 88.1% of patients with word recognition scores <44.2% qualified for cochlear implantation using private insurance criteria regardless of the sentence level testing material used for implant candidacy evaluation. CI 5 cochlear implant; HINT 5 Hearing in Noise Test; MWRT 5 monosyllabic word recognition test; WRS 5 word recognition score. DISCUSSION Cochlear implantation is an effective means of hearing rehabilitation in individuals with bilateral severe to profound deafness. Identifying patients who might meet audiological candidacy for cochlear implantation can be challenging for hearing health professionals using routine audiometric assessment. Although specific US Food & Drug Administration criteria exist for determining CI candidacy, these criteria are based on results of a formal CI evaluation, which is a more involved testing paradigm than that used routinely in the clinic. We were unable to identify data in the existing literature on this topic examining the predictive ability of routine audiometry for determining CI candidacy upon formal CI evaluation. We sought to evaluate the correlation of discrete findings on routine audiometric assessment with the outcome of CI assessment in a retrospective fashion. Overall we identified a number of findings on the audiograms of patients referred for CI evaluation that strongly correlated with a patient s performance on CI audiometric evaluation. The strongest correlations of pure-tone testing thresholds with SWDT were those of the low frequencies. Interestingly, the correlations of the MWRT with SWDT were strongest for HINT sentences, though strong correlations were found for all three types of sentence testing material utilized in this study. After establishing the presence of significant correlations between findings on routine audiometry with those of the SWDT, we sought to identify thresholds for CI candidacy using pure tones and the MWRT. We identified thresholds for both pure tones and the MWRT that identified with reasonable accuracy patients who would qualify for a CI upon formal audiometric assessment. These data identified threshold values on elements of routine office-based audiometry that can potentially be used to assist hearing health professionals in their decision of who to refer for a formal CI evaluation (Table II, Fig. 8). In general, if a patient is struggling with his or her current amplification and has audiometric thresholds below 75 db at 250 to 4,000 Hz, or performs poorly on the MWRT (<40%), then one can expect a high likelihood of meeting CI candidacy. 220

6 Fig. 8. Summary showing the pure-tone and MWRT threshold values that significantly correlated with meeting candidacy requirements upon cochlear implant evaluation. The percentile scores in the bottom row represent the number of patients in the study cohort testing at or below the audiometric measure thresholds shown who then qualified for cochlear implantation. For pure-tone measures, the threshold values were generated through correlation with sentence-level discrimination testing using AzBio in quiet conditions and assuming Medicare candidacy criteria (40% or worse on sentence level discrimination tests). For the MWRT, the threshold values were generated through correlation with sentence level discrimination testing regardless of the type of material used (AzBio in quiet, AzBio in noise, or HINT in quiet). CI 5 cochlear implant; HINT 5 Hearing in Noise Test; MWRT 5 monosyllabic word recognition test; WRS 5 word recognition score. Importantly, it should be noted that the majority of our analysis used the Medicare criteria of 40% correct or worse on sentence-level word discrimination testing in best-aided conditions to determine CI candidacy. Sentence level testing thresholds as high as 60% are used by many private insurers to determine CI candidacy. As such, our data regarding minimum thresholds of MWRT would be expected to potentially exclude some patients with private insurance who might be CI candidates. For example, in Figures 4 and 5, when comparing Medicare criteria to private insurance criteria, the predictive value of our data stays roughly the same (85.5% vs. 88.1%), whereas the MWRT threshold scores for reaching approval would be over 11 db higher (32.8 db vs db) for patients with private insurers. Clinicians should integrate this into their decision making regarding the appropriate hearing loss candidates to refer for CI evaluation. There are a number of limitations to this retrospective study. First, our study population was a preselected group of adult patients who had already been referred for a CI evaluation. Compared to the general population of patients with hearing loss, our preselected population of patients was thus presumably having significant enough problems with the use of hearing aids to warrant referral and evaluation for cochlear implantation. Therefore, any potential application of the pure-tone and MWRT minimum thresholds could be applied to aid in making clinical decisions for patients with hearing loss who are receiving suboptimal benefit from amplification. Another limitation to this study lies in our use of recorded voice rather than monitored live voice in presenting monosyllabic words. Although it is recommended to use recorded voice for testing, 15,16 it is currently not universally utilized in routine audiometric assessment. 17 In general, one would expect the MWRT values to be lower when using recorded voice than with live voice, but a recent study showed that the mode of presentation did not significantly impact the percent of words correct when testing normal-hearing and hearing-impaired listeners. 18 Despite the lack of a significant difference between live and recorded presentation modes in this study, the authors did advocate for the use of recorded voice presentations, as applied uniformly in our study, whenever possible. 18 In any event, our data would represent again a conservative estimate of potential CI candidacy given our use of recorded voice materials and should be interpreted by hearing health providers with this in mind. It is not always clear to healthcare professionals of different backgrounds (otolaryngologists, audiologists, hearing instrument specialists) when to refer individuals for a CI evaluation based upon routine audiometric testing. Because of this, patients who may benefit from a CI may not be referred for evaluation. A recent study out of the United Kingdom highlights this trend showing that fewer than half of the audiologists surveyed felt comfortable discussing CIs with their patients, and then actually making a referral for a CI evaluation. 12 A similar study surveying primary care physicians noted that although they were aware of the detrimental effect hearing loss has on quality of life, only 60% actually screened for hearing loss, and just over 25% of the physicians had actually made a referral for a CI evaluation. 13 Notably, both of these studies mentioned either a lack of training in referral criteria or an uncertainty of which patients were potential candidates as contributory. The present study attempts to clarify this process by showing that findings from routine audiometric assessment can be used to identify patients who have a high likelihood of qualifying for a CI upon formal evaluation. Application of the findings of this study to help guide appropriate referrals may provide an avenue for improving access to a CI for individuals struggling with hearing aids. Future prospective studies to evaluate the effectiveness of this approach will provide insight for further improvements toward achieving this goal. CONCLUSION Our study demonstrates that findings on routine office-based audiometry can be used to select patients with hearing loss who receive suboptimal benefit from amplification that have a high likelihood of meeting audiological candidacy for cochlear implantation (Table II, Fig. 8). In general, patients with low-frequency thresholds (250, 500, 1,000 Hz) greater than 75 db HL (250 2,000 Hz) or those with an MWRT score less than 40% using recorded voice-testing materials have a greater than 80% probability of meeting Medicare CI 221

7 candidacy requirements. These data may aid hearing health professionals in making appropriate referrals for CI evaluation thereby potentially improving access for those who would benefit. BIBLIOGRAPHY 1. Vermeire K, Brokx JPL, Wuyts FL, Cochet E, Hofkens A, Van de Heyning PH. Quality-of-life benefit from cochlear implantation in the elderly. Otol Neurotol 2005;26: Olze H, Szczepek AJ, Haupt H, Zirke N, Graebel S, Mazurek B. The impact of cochlear implantation on tinnitus, stress and quality of life in postlingually deafened patients. Audiol Neurootol 2012;17: Rembar S, Lind O, Arnesen H, Helvik AS. Effects of cochlear implants: a qualitative study. Cochlear Implants Int 2009;10: Ramos A, Guerra-Jimenez G, Rodriguez C, Borkoski S, Falcon JC, Perez D. Cochlear implants in adults over 60: a study of communicative benefits and the impact on quality of life. Cochlear Implants Int 2013; 14: Gaylor JM, Raman G, Chung M, et al. Cochlear implantation in adults: a systematic review and meta-analysis. JAMA Otolaryngol Head Neck Surg 2013;139: Budenz CL, Cosetti MK, Coelho DH, et al. The effects of cochlear implantation on speech perception in older adults. J Am Geriatr Soc 2011;59: Monteiro E, Shipp D, Chen J, Nedzelski J, Lin V. Cochlear implantation: a personal and societal economic perspective examining the effects of cochlear implantation on personal income. J Otolaryngol Head Neck Surg 2012;41(suppl 1):S43 S Cheng AK, Rubin HR, Powe NR, Mellon NK, Francis HW, Niparko JK. Cost-utility analysis of the cochlear implant in children. JAMA 2000; 284: Francis HW, Chee N, Yeagle J, Cheng A, Niparko JK. Impact of cochlear implants on the functional health status of older adults. Laryngoscope 2002;112(8 pt 1): Semenov YR, Martinez-Monedero R, Niparko JK. Cochlear implants: clinical and societal outcomes. Otolaryngol Clin N Am 2012;45: Wyatt JR, Niparko JK, Rothman M, delissovoy G. Cost utility of the multichannel cochlear implants in 258 profoundly deaf individuals. Laryngoscope 1996;106: Chundu S, Buhagiar R. Audiologists knowledge of cochlear implants and their related referrals to the cochlear implant centre: pilot study findings from UK. Cochlear Implants Int 2013;14: Cohen SM, Labadie RF, Haynes DS. Primary care approach to hearing loss: the hidden disability. Ear Nose Throat J 2005;84:26,29 31, Sorkin DL. Cochlear implantation in the world s largest medical device market: utilization and awareness of cochlear implants in the United States. Cochlear Implants Int 2013;14(suppl 1):S4 S Wiley TL, Stoppenbach DT, Feldhake LJ, Moss KA, Thordardottir ET. Audiologic practices: what is popular versus what is supported by the evidence. Am J Audiol 1995;4: Mullennix JW, Pisoni DB, Martin CS. Some effects of talker variability on spoken word recognition. J Acoust Soc Am 1989;85: Martin FN, Champlin CA, Chambers JA. Seventh survey of audiometric practices in the United States. J Am Acad Audiol 1998;9: Mendel LL, Owen SR. A study of recorded versus live voice word recognition. Int J Audiol 2011;50: