Abstract
Hearing loss is common among Veterans, and extensive hearing care resources are prioritized within the Veterans Administration (VA). Severe hearing loss poses unique communication challenges with speech understanding that may not be overcome with amplification. We analyzed data from the VA Audiometric Repository between 2005 and 2017 and the relationship between hearing loss severity with speech recognition scores. We hypothesized that a significant subset of Veterans with severe or worse hearing loss would have poor unaided speech perception outcomes even with adequate audibility. Sociodemographic characteristics and comorbidities were compiled using electronic medical records as was self-report measures of hearing disability. We identified a cohort of 137,500 unique Veterans with 232,789 audiograms demonstrating bilateral severe or worse hearing loss (four-frequency PTA > 70 dB HL). The median (IQR; range) age of Veterans at their first audiogram with severe or worse hearing loss was 81 years (74 to 87; 21–90+), and a majority were male (136,087 [99%]) and non-Hispanic white (107,798 [78.4%]). Among those with bilateral severe or worse hearing loss, 41,901 (30.5%) also had poor speech recognition scores (<50% words), with greater hearing loss severity correlating with worse speech perception. We observed variability in speech perception abilities in those with moderate-severe and greater levels of hearing loss who may derive limited benefit from amplification. Veterans with communication challenges may warrant alternative approaches and treatment strategies such as cochlear implants to support communication needs.
Keywords: hearing loss, veteran health, audiometry, age-related hearing loss
Introduction
Hearing loss impairs communication, limits function, and negatively affects quality of life. With untreated or undertreated hearing loss, communication difficulties pervade personal, professional, and healthcare spheres (Mulrow et al., 1990). Hearing loss is one of the most common service-connected health conditions among United States Veterans (Wells et al., 2015). In addition to cumulative effects of harmful noise exposure from military service (Beck, 2010; Centers for Disease & Prevention, 2011; Cooper et al., 2014), further loss of hearing occurs with advancing age (Lin et al., 2011). Presbycusis or age-related hearing loss affects more than 25% of adults over 60 years of age and more than 80% of adults over 80 years of age (Goman & Lin, 2016).
As of 2020, more than 1.3 million Veterans were receiving disability compensation for hearing loss (Veterans Benefits Administration, 2021). In 2022, the Veterans Affairs (VA) spent $430 million for 1,000,000 hearing aids (Historical VA Hearing Aids Procurement Distribution Summaries) Although the provision of hearing care is costly, it enables Veterans access to high quality hearing aids at no cost (VHA Directive 2008-070). Although hearing aids are a first line treatment, appropriateness of the rehabilitation strategy differs based on the degree of hearing loss. Hearing aids improve audibility and are beneficial for mild–moderate hearing loss. The effectiveness of amplification may be limited in Veterans with greater hearing loss because the ability to hear a sound is not sufficient for speech understanding. Severe hearing loss poses unique rehabilitative challenges compared to lesser degrees of hearing loss, and individuals with such loss may rely more on visual cues (Atcherson et al., 2017) and cognitive repair mechanisms (Lyxell et al., 2003).
Presbycusis and noise exposure both lead to loss of sensory hair cells in regions of the cochlea that respond to high frequency sounds. This loss leads to hearing decline and difficulty with speech understanding. Mechanistically, this may result from outer or inner hair cell damage, leading to broadened auditory filters and reduced frequency selectivity, and so-called “dead regions” impairing speech understanding (Faulkner et al., 1990). More simply, in conversational speech, different speech sounds occur at different frequencies. Speech recognition particularly depends on perceiving high frequency sounds, and once a critical number of hair cells are lost from the basal region of the cochlea, amplification with hearing aids may not restore the ability to recognize these sounds.
Clarity in speech perception is readily assessed during a comprehensive audiometric evaluation and is denoted by the word recognition score. This score is derived from the percent correct of individual word lists presented at a sound level that maximizes performance and avoids loudness discomfort. Although there is an established association between severity of hearing loss and poorer speech recognition scores (Carhart & Porter, 1971; Yoshioka & Thornton, 1980), there is wide inter-individual variability. This variability may be related to aging effects on auditory processing (Pichora-Fuller & Souza, 2003), in addition to the etiology of hearing loss, and duration of deafness (Pichora-Fuller & Souza, 2003).
The epidemiology and prevalence of Veterans with severe hearing loss have received little attention; moreover, the impact of severe hearing loss on speech perception abilities for Veterans has not been well-characterized. An indeterminate number of Veterans have difficulty understanding speech even with hearing aids. We lack an appreciation of how many Veterans have severe hearing loss and what proportion of those Veterans will receive little benefit from hearing aids, thus requiring other management strategies.
Comprehensive diagnostic audiology and hearing rehabilitation services are available to all Veterans, increasing the likelihood that Veterans seek these services within the VA system. VA audiometric data are centralized in a repository containing all diagnostic audiogram elements as well as self-reported hearing assessments. This data set has been used previously to examine particular cohorts such as Veterans who had normal audiograms (Billings et al., 2018) or Veterans with configurations suggestive of prior noise exposure (Wilson & McArdle, 2013), but none focused on severe hearing loss. In this paper, we use the VA audiometric repository to provide a descriptive analysis of Veterans with severe hearing loss who utilized VA audiology services between 2005 and 2017. By correlating hearing loss severity with speech scores, we can identify opportunities for improvement in health services by articulating the proportion of Veterans with variable benefit from audibility gain and therefore in need of direct assessment of their outcomes with hearing aids, thus requiring other strategies to support communication needs.
Methods
Data Acquisition
We obtained data from the Denver Logistics Center (DLC) of the U.S. Department of Veterans Affairs Office of Procurement, Acquisition and Logistics (OPAL). The DLC manages the National VA Audiometric Repository including behavioral pure tone thresholds and speech perception data. All audiometric data from VA facilities were uploaded to the repository between 2005 and 2017. At the time of data extraction, the Audiometric Repository contained a total of 5,075,363 audiograms from 1,938,545 unique Veterans.
As outlined in other studies utilizing VA electronic databases (Dillard et al., 2020; Saunders et al., 2021), substantial data cleaning was required prior to analysis. Institutional Review Board approval was obtained through the Department of Veterans Affairs New York Harbor Healthcare System (NYHHS IRB ID 01694).
Cohort Identification and Hearing Loss Categories
We extracted data for all VA audiograms obtained during the study period with air conduction (AC) averages (1–8 kHz) greater than or equal to 70 dB HL for both ears. We then computed the four-frequency pure tone average (4FPTA) across 0.5, 1, 2, and 4 kHz to generate a single index of hearing loss severity (Lin & Reed, 2021). We excluded audiograms that had available thresholds for fewer than three of the four frequencies. Only Veterans with at least one audiogram that had severe or worse bilateral hearing loss (4FPTA ≥ 70 dB HL) at any evaluation between 2005 and 2017 were included in the analysis. To observe longitudinal changes, we extracted all available audiograms for any Veteran, meeting the inclusion criteria bilaterally—both prior to and subsequent to the “qualifying” date when severe or worse hearing loss was first recorded. We also extracted the suprathreshold measure of speech perception based on a monosyllabic word test in quiet, most commonly the Central Institute of the Deaf W-22 (CID W-22) or the Northwestern University-6 word list (NU-6). This variable is expressed as the word recognition score (WRS) for each ear at the time of each audiogram. WRS for each ear at earlier time points was included when available.
Veterans were further classified into one of three groups—“Severe bilaterally” (4FPTA between 70 and 89 dB HL for both ears), “Profound bilaterally” (4FPTA ≥ 90 dB HL for both ears), and “1 Severe ear/1 Profound ear.” To estimate how often Veterans with severe hearing loss have difficulty with speech perception despite what we infer was adequate audibility, we examined the correlation between hearing severity thresholds (4FPTA) and the suprathreshold measure of speech perception (WRS).
Mixed hearing loss was assessed using bone-conduction (BC) averages, a variable within the Repository, which represents an average of all audiometric thresholds recorded between 1 and 6 kHz for BC thresholds. We defined a conductive component as meaningful if there was a 30 dB air–bone gap. Only ∼5% of records (54,611/990,592) included a BC average, and so we could only calculate the AC − BC difference for this subset of records. Among these, 645 records had a ≥30 dB air–bone gap for the left ear alone; 643 for the right ear alone; and 239 for both ears (affecting 1,286 unique individuals, or 2.4% of this population). Given this, we chose not to exclude these Veterans from the cohort.
We identified the subset of Veterans who were fitted with hearing aids who also completed the patient-reported outcome measure used most widely in the VA, the International Outcome Inventory for Hearing Aids (IOI-HA), which is also uploaded to the Repository. We paired IOI-HA responses with audiograms performed within 1 year (±365 days) of survey completion to compare the measured and self-reported degree and impact of hearing loss. This is currently the only outcome measure for hearing treatment in the VA Audiometric Repository.
Data Management
There were 1,980 exact duplicate audiograms, of which only one record was kept for each. There were also nonidentical duplicate audiograms that were completed on the same day for a given Veteran (n = 1,775 audiograms). In these cases, we used the better hearing or lowest 4FPTA from that day.
Nonnumeric values were managed using several data cleaning methods. This included approaches used in prior literature utilizing the VA Audiometric Repository (Saunders et al., 2021) as well as our team's own strategies given our particular focus on those with severe hearing loss. This was important as nonnumeric values may be used in audiometry to denote responses at the limits of testing abilities, which would be particularly relevant to those with severe or worse hearing loss.
To address symbols, presumably indicating that no response was obtained at the upper limits of the audiometer (i.e., +, *), thresholds ≥90 dB HL with these nonnumeric values were converted to 120 dB HL (Saunders et al., 2021). Thresholds <90 dB HL with nonnumeric values present were converted to missing. Additionally, some responses for word recognition scores included the nonnumeric responses “could not test” (CNT) and “did not test” (DNT). If a WRS was listed as CNT or DNT for an individual with 4FPTA ≥100 dB HL (indicating profound hearing loss) or WRS ≤40% (indicating poor speech recognition) at any prior audiogram on record, the WRS was flagged as <50%, on the assumption that hearing was too poor to assess speech recognition. If the WRS was listed as CNT/DNT for an individual with 4FPTA <100 dB HL and all prior WRS >40%, the response was treated as missing since the reason for listing CNT/DNT could not be determined. Finally, if the WRS was missing for a Veteran who had a recorded WRS from the previous 12 months, that score was carried forward to the visit with a missing WRS. Missing WRS data for Veterans who previously had WRS <40% were flagged as <50%, and missing WRS for all other circumstances remained classified as missing.
Patient-Level Measures From Electronic Medical Records
We derived age, gender, self-reported race and ethnicity, urban or rural status of residence, and VA “enrollment priority” based on military service and income from VA enrollment data for the fiscal year ([FY] VA fiscal years begin on 1 October and end on 30 September of the following year) of an individual's first audiogram indicating severe or worse hearing loss. Veterans are assigned to one of eight priority groups upon VA enrollment with lower numbers generally corresponding to a higher priority. We categorized priority status into three groups, as others have done (Radomski et al., 2022), based on the generosity of VA health benefits: service-connected disability present, low income, and no service-connected disability/low income. Comparisons were made between Veterans’ first and most recent audiogram on record.
We summarized the demographics of Veterans with severe or worse hearing loss and considered subgroups of those based on WRS with regard to age, race, rurality, and comorbidities. We identified the number of chronic medical conditions using the International Classification of Diseases, Ninth and Tenth Revisions (ICD-9 and ICD-10) diagnosis codes for 30 Elixhauser conditions (Quan et al., 2005). We coded conditions as present if they occurred at either an outpatient or an inpatient visit in the 5 years prior to an individual's first audiogram with severe or worse hearing loss.
Statistical Analyses
Prior to deriving our cohort of interest, we visualized the association between word recognition score and 4FPTA for all available audiograms using boxplots; specifically, for all ears with available information for both WRS and 4FPTA (1,645,747 ears, representing 463,787 unique individuals), we created boxplots of WRS by decile of 4FPTA (i.e., <10, 10–20, 20–30, etc.), with the final decile representing those with 4FPTA of 100 dB or more.
Among the cohort of interest (n = 137,500), we first calculated descriptive statistics for age, number of audiograms, hearing severity, and speech perception for Veterans at the time of their first audiogram with severe or worse hearing loss (4FPTA ≥ 70 dB HL). For those with multiple audiograms following their initial audiogram with bilateral severe or worse hearing loss, WRS was grouped by degree of hearing loss. We also tabulated descriptive characteristics by whether an individual with severe hearing loss also had a WRS <50% (“Severe or worse HL and poor WRS”) or not (“Severe or worse HL but better-than-poor WRS”). Groups were compared using t-tests for continuous variables and Chi-square or Fisher's exact test for categorical variables. All tests were two-tailed, with a Bonferroni-adjusted p-value of 0.006 to account for multiple comparisons (0.05/9 comparisons = 0.006).
To examine the association between self-rated hearing difficulty from the International Outcome Inventory for Hearing Aids (IOI-HA) survey and measured hearing loss from audiometric data, we used the Pearson correlation coefficient and also depicted the overlap between hearing loss categorizations using a heat map. Of note, only audiogram-IOI survey pairs within 1 year (±365 days) of each other were used. All data cleaning and analyses were performed with SAS Enterprise Guide version 8.3 and SAS 9.4 (SAS Institute Inc).
Results
Figure 1 displays the association between WRS and 4FPTA for all available audiograms using boxplots for each decile of 4FPTA. With increasing 4FPTA, we noted a near-linear downward trend in both mean and median WRS. Despite the strong correlation between WRS and 4FPTA, we did notice a great deal of variability in WRS across all deciles, which is perhaps best illustrated by the outliers present in the distributions ≤40 dB (i.e., those with a 4FPTA between 20 and 40 dB and a WRS indicating poor or very poor speech recognition). The median (IQR) WRS for those with 4FPTA < 10 dB was 98 (6), while the median (IQR) for those with 4FPTA ≥ 100 dB was 8 (32). Among those ears with a 4FPTA ≥ 70 dB, the upper quartile of the distribution fell within the WRS region of “very poor” (WRS 0–52%) to “fair” (WRS 66–76%).
Figure 1.
Relationship between severity of hearing loss (4FPTA) and speech recognition (WRS). Boxplots of WRS by decile of 4FPTA for the total data set (1,645,747 ears, representing 463,787 unique individuals) with the final decile representing those with 4FPTA of 100 dB or more. Red dotted reference lines are included to denote established speech recognition category cutoffs. Background shading is used to highlight the distinct speech recognition categories of “very poor” (0–52%), “poor” (54–64%), “fair” (66–76%), “good” (78–88%), and “excellent” (≥90%) with darker shades of blue corresponding to worse speech recognition. Of note, though the primary interest of this paper is to identify and characterize individuals with severe or worse hearing loss, we have included all audiograms—including those considered better-than-severe—for this visualization.
Figure 2 details the number of unique Veterans and audiograms included in each step to derive our cohort of interest. There were 1,010,815 audiograms with 70 dB HL or poorer from 1 to 8 kHz, representing 465,480 unique Veterans. Audiograms that did not have thresholds for both ears for at least three of the four frequencies from the 4FPTA (3,245, 0.3%) or did not have 4FPTA poorer than 70 dB HL bilaterally were excluded (643,387, 64.6%). This yielded a cohort of 137,500 unique Veterans and a total 347,205 audiograms. Of these, 232,789 audiograms demonstrated bilateral severe or worse hearing loss.
Figure 2.
Data cleaning and management processes to derive the cohort of interest. DLC = Denver Logistics Center; 4FPTA = four-frequency pure tone average; WRS = word recognition score; kHz = kilohertz; dB= decibel.
The median age of Veterans with severe or worse hearing loss bilaterally was 81 years for the first audiogram (IQR, 74 to 87; range 21–90+) and 83 years (IQR, 75 to 88; range 21–90+) for their most recent audiogram (Table 1). This cohort had a median (IQR; range) of two audiograms (1 to 3; 1–27) during this study period. The mean number of audiograms increased with hearing loss suggesting that more severely impaired individuals continued to seek audiological care for their disability. From the cohort of 137,500 Veterans who had 4FPTA greater than 70 dB HL bilaterally at any evaluation between 2005 and 2017, 89,719 (65.3%) had two or more audiograms; of those, 41,150 (45.9%) had severe hearing loss for their first VA audiogram, while 77,545 (86.4%) had severe hearing loss at their most recent audiogram. The mean (SD) interval between the first and most recent audiograms for this subset was 3.5 (3.5) years (median 3, IQR 0 to 6, range 0–12). The mean (SD) duration of follow-up with audiology among those with bilateral severe or worse hearing loss was 1.8 (2.7) years (median 0, IQR 0 to 3, range 0–12).
Table 1.
Age, Number of Audiograms, 4FPTA, and Speech Recognition (WRS) at First Audiogram With Severe or Worse HL Among Those With 1 + Audiogram With Bilateral Severe or Worse Hearing Loss (N = 137,500); by Severity of Hearing Loss.
| Overall | Severe HL bilaterally | 1 Profound/1 Severe | Profound HL bilaterally | |
|---|---|---|---|---|
| N = 137,500 | n = 101,855 | n = 24,853 | n = 10,792 | |
| Age | ||||
| Age at first audiogram with severe or worse HL, years | 79.7 (9.7) | 80 (9.5) | 79.2 (10.0) | 77.8 (10.7) |
| Age at first audiogram with bilateral severe or worse HL and WRS < 50%, years | 81.1 (9.3) | 82 (8.7) | 80.5 (9.5) | 78.3 (10.6) |
| Age at last audiogram, years a | 81.3 (9.6) | 81.7 (9.4) | 80.8 (9.9) | 79.5 (10.4) |
| Number of audiograms | ||||
| Audiograms during 2005–2017 | 2.5 (1.7) | 2.6 (1.7) | 2.5 (1.7) | 2.1 (1.4) |
| Severe or worse HL audiograms | 1.7 (1.2) | 1.7 (1.1) | 1.7 (1.2) | 1.8 (1.2) |
| Severe or worse HL audiograms + WRS < 50% | 0.6 (0.9) | 0.5 (0.9) | 0.6 (1.0) | 1.2 (1.2) |
| 4FPTA | ||||
| Left ear | 80.8 (11.0) | 76.4 (5.1) | 89.4 (13.4) | 101.9 (9.4) |
| Right ear | 80.5 (11.0) | 76.1 (5.1) | 89.1 (13.4) | 101.7 (9.4) |
| Severe ear | 78.5 (5.9) | – | 78.5 (5.9) | – |
| Profound ear | 100.0 (9.8) | – | 100.0 (9.8) | – |
| Speech recognition scores | ||||
| Continuous WRS | ||||
| Left ear | 49.5 (25.1) | 52.0 (23.5) | 43.6 (28.2) | 29.9 (26.3) |
| Right ear | 50.5 (25.3) | 53.1 (23.7) | 44.9 (28.6) | 30.4 (26.5) |
| Bilateral average b | 52.1 (25.8) | 54.7 (24.0) | 49.6 (28.2) | 28.5 (27.2) |
| Severe ear | 31.3 (26.5) | – | 31.3 (26.5) | – |
| Profound ear | 54.1 (25.8) | – | 54.1 (25.8) | – |
| Categorical WRS (%N) | ||||
| Both <50% | 41,901 (30.5) | 27,868 (27.4) | 7,626 (30.7) | 6,407 (59.4) |
| Both ≥50% | 44,025 (32) | 39,591 (38.9) | 3,529 (14.2) | 905 (8.4) |
| 1 ear ≥50%/ 1 ear <50% | 27,246 (19.8) | 17,881 (17.6) | 8,187 (32.9) | 1,178 (10.9) |
| 1 ear miss/1 ear <50% | 4,390 (3.2) | 994 (1) | 2,148 (8.6) | 1,248 (11.6) |
| 1 ear miss/1 ear ≥50% | 959 (0.7) | 465 (0.5) | 469 (1.9) | 25 (0.2) |
| Missing in both | 18,979 (13.8) | 15,056 (14.8) | 2,894 (11.6) | 1,029 (9.5) |
Note: all values presented as mean (SD) or n (%). HL = hearing loss; WRS = word recognition score; 4FPTA = four-frequency pure tone average.
Calculated for those individuals with 1+ audiogram following their first audiogram with bilateral severe or worse hearing loss (n = 62,696; severe bilaterally 46,009; 1 profound/1 severe 11,654; profound bilaterally 5,033).
Bilateral average calculated in the case of nonmissing WRS values for both ears and the difference between ears was <20%.
Of the 137,500 Veterans with bilateral severe hearing loss, 54,431 (39.6%) also had WRS <50% bilaterally, as shown in Figure 1. While less than 1% of pure tone data were missing from the original DLC dataset, WRS data had a greater percentage of missing values with increasing hearing loss prior to implementing cleaning rules (Appendix A). We summarize the WRS value types for all records pre- and postcleaning rules based on the initial data type (e.g., numeric WRS, CNT, DNT, missing) in Appendix B. The percentage of individuals with WRS <50% bilaterally ranged from 27.4% among those with severe HL in both ears, 30.7% among those with one profound and one severe ear, and 59.4% among those with profound hearing loss bilaterally (Cochran Armitage test for trend p < .0001). We considered longitudinal changes in speech perception abilities from the time of severe hearing loss diagnosis both considering scores as a continuous variable and categorically what is classified as poor word recognition abilities (bilateral WRS <50%) (Table 2).
Table 2.
Speech Recognition (WRS) at First Audiogram With Severe or Worse HL and Last Audiogram Among Those With 1 + Audiogram Following Their First Audiogram With Bilateral Severe or Worse Hearing Loss; by Hearing Loss Group.
| Overall | Severe HL bilaterally | 1 Profound/1 Severe | Profound HL bilaterally | |||||
|---|---|---|---|---|---|---|---|---|
| First | Last | First | Last | First | Last | First | Last | |
| N | 62,696 | 62,696 | 46,009 | 46,009 | 11,654 | 11,654 | 5,033 | 5,033 |
| Time between first severe or worse HL audiogram and last, years a | 4.0 (2.7) | 4.0 (2.7) | 3.9 (2.8) | 4.1 (2.9) | ||||
| Continuous WRS | ||||||||
| Left ear | 51.3 (25.2) | 47.1 (26.0) | 54.0 (23.6) | 49.0 (24.6) | 45.5 (28.2) | 43.5 (28.6) | 32.5 (26.5) | 32.3 (29.0) |
| Right ear | 52.3 (25.5) | 48.2 (26.2) | 55.0 (23.7) | 50.2 (24.8) | 46.4 (28.7) | 44.3 (29.0) | 32.9 (26.8) | 33.0 (29.3) |
| Bilateral average a | 54.4 (25.7) | 491. (27.1) | 57.0 (23.8) | 51.1 (25.6) | 51.5 (28.1) | 47.9 (28.9) | 31.6 (27.7) | 31.4 (30.2) |
| Severe ear | 55.8 (25.6) | 49.0 (25.4) | – | – | 55.8 (25.6) | 52.0 (25.3) | – | – |
| Profound ear | 33.3 (26.9) | 29.3 (24.6) | – | – | 33.3 (26.9) | 25.3 (24.5) | – | – |
| Categorical WRS (%N) | ||||||||
| Both <50% | 17,288 (27.6) | 25,213 (40.2) | 11,191 (24.3) | 16,858 (36.6) | 3,296 (28.3) | 4,890 (42) | 2,801 (55.7) | 3,465 (68.8) |
| Both ≥50% | 21,231 (33.9) | 18,310 (29.2) | 18,841 (41) | 15,801 (34.3) | 1,890 (16.2) | 1,909 (16.4) | 500 (9.9) | 600 (11.9) |
| 1 ear ≥50%/1 ear <50% | 12,327 (19.7) | 12,466 (19.9) | 7,930 (17.2) | 8,352 (18.2) | 3,807 (32.7) | 3,562 (30.6) | 590 (11.7) | 552 (11) |
| 1 ear miss/1 ear <50% | 1,918 (3.1) | 2,315 (3.7) | 388 (0.8) | 1,283 (2.8) | 948 (8.1) | 772 (6.6) | 582 (11.6) | 260 (5.2) |
| 1 ear miss/1 ear ≥50% | 461 (0.7) | 217 (0.3) | 215 (0.5) | 140 (0.3) | 232 (2) | 69 (0.6) | 14 (0.3) | 8 (0.2) |
| Missing in both | 9,471 (15.1) | 4,175 (6.7) | 7,444 (16.2) | 3,575 (7.8) | 1,481 (12.7) | 452 (3.9) | 546 (10.8) | 148 (2.9) |
Note. “First” refers to the first audiogram meeting bilateral severe or worse hearing loss. “Last” refers to a person's final audiogram. All values presented as mean (SD) or n (%) unless otherwise indicated.
HL = hearing loss; WRS = word recognition score.
Bilateral average calculated in the case of non-missing WRS values for both ears and the difference between ears was <20%.
We considered sociodemographic factors, measures of rurality, and comorbidities as conveyed by the Elixhauser index among those Veterans with severe or worse hearing loss and then separately among those who did and did not have poor word recognition abilities (Table 3). There were no significant differences, including notable comorbidities. These were similarly distributed among Veterans with severe or worse hearing loss who did and did not have poor word recognition abilities (Table 3).
Table 3.
Patient Characteristics at Incident Audiogram Meeting Criteria for Those With Bilateral Severe or Worse Hearing Loss (4FPTA) and Better Than “Poor” Word Recognition Score (WRS) versus. Those With Bilateral Severe or Worse Hearing Loss and “Poor” Word Recognition Score (Less Than 50% in Both Ears).
| N | All with severe or worse HL bilaterally | Severe or worse HL but better-than poor WRS | Severe or worse HL and poor WRS | p-value |
|---|---|---|---|---|
| 137,500 | 95,599 | 41,901 | ||
| Mean age in years (SD) | 79.7 (9.7) | 79.0 (9.8) | 81.1 (9.3) | <.0001 |
| Age category, n (%) | <.0001 | |||
| <=50 | 924 (0.7) | 710 (0.7) | 214 (0.5) | |
| 51–70 | 23,906 (17.4) | 18,206 (19) | 5700 (13.6) | |
| 71–85 | 63,785 (46.4) | 45,399 (47.5) | 18,386 (43.9) | |
| 85+ | 48,885 (35.6) | 31,284 (32.7) | 17,601 (42) | |
| % male | 136,087 (99) | 94,603 (99) | 41,484 (99) | .21 |
| Race and ethnicity, n (%) | <.0001 | |||
| NH White | 107,798 (78.4) | 75,146 (78.6) | 32,652 (77.9) | |
| NH Black or African American | 5132 (3.7) | 3578 (3.7) | 1554 (3.7) | |
| NH Asian | 676 (0.5) | 402 (0.4) | 274 (0.7) | |
| Hispanic | 3780 (2.7) | 2561 (2.7) | 1219 (2.9) | |
| NH Other | 4008 (2.9) | 2817 (2.9) | 1191 (2.8) | |
| Missing | 16,106 (11.7) | 11,095 (11.6) | 5011 (12) | |
| Rurality, n (%) | .18 | |||
| Urban | 75,501 (54.9) | 52,191 (54.6) | 23,310 (55.6) | |
| Rural or highly rural | 58,109 (42.3) | 40,368 (42.2) | 17,741 (42.3) | |
| Missing | 3890 (2.8) | 3040 (3.2) | 850 (2) | |
| VA priority group, n (%) | <.0001 | |||
| 1–4 | 69,701 (50.7) | 49,647 (51.9) | 20,054 (47.9) | |
| 5 | 28,639 (20.8) | 18,958 (19.8) | 9681 (23.1) | |
| 6–8 | 35,622 (25.9) | 24,196 (25.3) | 11,426 (27.3) | |
| Hearing loss severity, n (%) | <.0001 | |||
| Severe in both | 101,855 (74.1) | 73,987 (77.4) | 27,868 (66.5) | |
| 1 Profound, 1 Severe | 24,853 (18.1) | 17,227 (18) | 7626 (18.2) | |
| Profound in both | 10,792 (7.8) | 4385 (4.6) | 6407 (15.3) | |
| No. of Elixhauser conditions, mean (SD) | 3.0 (2.3) | 3.1 (2.4) | 3.0 (2.3) | <.0001 |
| Total Elixhauser comorbidities, n (%) | <.0001 | |||
| None | 15,213 (11.1) | 10,478 (11) | 4735 (11.3) | |
| 1 | 22,023 (16) | 14,988 (15.7) | 7035 (16.8) | |
| 2 | 26,358 (19.2) | 18,184 (19) | 8174 (19.5) | |
| 3+ | 69,049 (50.2) | 48,925 (51.2) | 20,124 (48) | |
| Most prevalent Elixhauser comorbidities, n (%) | ||||
| Hypertension | ||||
| Uncomplicated hypertension | 86,086 (62.6) | 60,172 (62.9) | 25,914 (61.8) | .63 |
| Complicated hypertension | 10,872 (7.9) | 7610 (8) | 3262 (7.8) | .26 |
| Chronic pulmonary disease | 34,120 (24.8) | 24,114 (25.2) | 10,006 (23.9) | <.0001 |
| Depression | 30,019 (21.8) | 21,764 (22.8) | 8255 (19.7) | <.0001 |
| Solid tumor without metastasis | 23,627 (17.2) | 16,599 (17.4) | 7028 (16.8) | .09 |
| Diabetes | ||||
| Diabetes, uncomplicated | 23,033 (16.8) | 16,435 (17.2) | 6598 (15.7) | <.0001 |
| Diabetes, complicated | 15,605 (11.3) | 11,312 (11.8) | 4293 (10.2) | <.0001 |
| Obesity | 22,363 (16.3) | 16,485 (17.2) | 5878 (14) | <.0001 |
| Peripheral vascular disease | 19,562 (14.2) | 13,681 (14.3) | 5881 (14) | .63 |
Characteristics were evaluated using t-tests for continuous variables and Chi-square or Fisher's exact test for categorical variables.
Veterans are assigned to one of eight priority groups at VA enrollment based on service-connected illnesses, era of service, and socioeconomic status determined by means testing. Priority group determines the level of copayment. Priority groups are categorized for presentation here based on similarity of copays between groups.
HL = hearing loss; NH = non-Hispanic; VA = US Department of Veterans Affairs; WRS = word recognition score; 4FPTA = the four-frequency pure tone average.
In total, 88,523 Veterans had an IOI-HA survey of patient-reported outcomes linked to an audiogram within 1 year of survey completion (Table 4). Among our cohort, 25,900 (18.8%) individuals completed an IOI-HA survey within 1 year of an available audiogram. Self-reported hearing loss increased as 4FPTA increased (r = .27, p < .0001), though there was considerable variability; this variability is highlighted in the heat map of self-reported hearing loss versus hearing loss for the closest audiogram. For example, among those with better-than-severe hearing loss bilaterally, 31% of individuals self-reported as having ‘severe’ hearing loss. Despite the variability, there was an increase in the percentage of individuals who self-reported as having ‘severe’ hearing loss with increasing hearing loss as assessed using their closest audiogram (63% among those with severe in both ears, 73% among those with one profound and one severe ear, and 82% among those with profound hearing loss bilaterally).
Table 4.
Measured Hearing Loss Category from Audiogram Closest to International Outcome Inventory for Hearing Aids (IOI-HA) Survey Completion a versus Self-Reported Degree of Hearing Loss from IOI-HA.
| IOI-HA self-reported hearing difficulty | ||||||
|---|---|---|---|---|---|---|
| None | Mild | Moderate | Moderately severe | Severe | Total | |
| Better-than-severe HL in both ears | 1.0% | 3.7% | 21.8% | 42.4% | 31.1% | 69,688 |
| Severe HL in both ears | 1.2% | 1.8% | 8.0% | 25.6% | 63.3% | 13,723 |
| 1 Profound HL ear, 1 Severe HL ear | 1.3% | 1.5% | 4.6% | 19.1% | 73.3% | 3,447 |
| Profound HL in both ears | 2.2% | 0.8% | 3.0% | 12.0% | 81.9% | 1,665 |
| Total | 1.1% | 3.3% | 18.6% | 38.3% | 38.7% | 88,523 |
Note. Percentages represent row percentages. HL = hearing loss.
Only audiogram-IOI survey pairs within 1 year (±365 days) of each other were used.
Table cells are colored from light to dark red with darker shades of red corresponding to higher percentages.
Discussion
Nearly 50% of U.S. Veterans are older than age 65 as of 2020 (Wang et al., 2021); increased age is positively correlated with prevalence of severe hearing loss (Groenewold et al., 2011) and negatively correlated with speech recognition scores (Hoppe et al., 2016). Speech recognition declines after age 60 (Jerger, 1973) and deteriorates by up to 25% across the lifespan (Hoppe et al., 2021). Speech recognition in quiet significantly declines after age 80 years (Hoppe et al., 2016) even when the audiogram remains normal, possibly related to reduced information-carrying capacity (Hoppe et al., 2016) among older individuals.
This study is the first to specifically study patterns of severe hearing from the expansive Audiometric Repository of Veterans, a unique data set recording comprehensive audiometric assessments. While we observed a wide range of speech recognition abilities for Veterans with severe hearing loss in both ears, consistent with prior findings, speech understanding worsened with increasing hearing loss (Jerger, 1973; Souza et al., 2018). Our current study of Veterans with severe hearing loss is unique given the size of the dataset and the consistent test battery applied based on protocols within the VA system. This is unique among large datasets; usually non-VA patients might be seen at a wider variety of practices with test batteries ranging from detailed speech testing to no speech testing. This allowed examination of the correlation between hearing loss severity and speech recognition scores which highlights the variability in unaided speech outcomes at different hearing thresholds (Figure 2). We expect aided scores with fitted amplification would also demonstrate marked variability, although not necessarily in a predictable way. This raises concern for potential of quality-of-care gap in care if such individuals are not being specifically assessed for outcomes with amplification to ensure tailored treatment is offered to overcome the deficit in speech understanding.
We observed a bias toward increased missingness in speech perception outcome variables associated in Veterans with worse hearing on audiometry. This may reflect the perceived futility of assessing speech understanding in a population with this level of hearing loss, and as such, the frequency of missingness may be a surrogate marker for a lack of speech recognition abilities. The increased frequency of missing speech recognition scores in those with worse levels of hearing sensitivity (4PTA) reflects our clinical observation that speech testing in such patients may not be possible or anticipated as impossible by the examiner. For this reason, discarding these missing data would bias relevant findings from our study.
Previous studies have looked at WRS as a function of age and gender in addition to worsening pure tone audiogram; however, the sample sizes were not nearly as large as is available from this Repository, evaluated word recognition performance in quiet for 3,189 adults ranging from 48 to 92 years of age (Dubno et al., 1997; Wiley et al., 1998). WRS was worse for older age groups and worse for men than for women. One other study with a sample size of only 129 individuals found significant declines in word recognition scores for men as age increased (Dubno et al., 1997). Humes and Roberts (1990) found the primary determinant of speech recognition performance for older adults to be threshold elevation. Among the 36 participants in the study, older patients with a more severe hearing loss scored lower on speech recognition testing (Humes & Roberts, 1990).
Given our focus on patients with severe or worse sensorineural hearing loss, we considered excluding patients with mixed patterns from a significant conductive component to the hearing loss, which may confound the relationship between pure tone thresholds and speech perception. We elected not to exclude Veterans on this basis given its relative infrequency in the database while acknowledging that these individuals may have better word recognition scores than the PTA would suggest since conductive hearing loss can be overcome by increasing the presenting volume level.
The approach to missingness used here eliminated 347,205 records, which constitute 34% of all observations. The remaining data represents the largest sample of its kind. This database can be integrated with data from the Veterans’ electronic health record (EHR), demographics, and other EHR-derived data (e.g., comorbidities, audiology clinic visits) and linked to the hearing data studies already underway to provide valuable insights related to Veteran characteristics and the longitudinal management of severe hearing loss (Saunders et al., 2021).
As the variability in these data highlights, WRS for moderate–severe hearing losses is not well predicted by the four-frequency pure tone average, and this is true for anticipating hearing aid benefit in this population as well (McRackan et al., 2016; Souza et al., 2018). There may be differences in scores for speech via headphones through an audiometer at the patient's maximal comfort level (“unaided”) and speech presented at conversational levels with use of hearing aids for amplification (“aided”). Among those with severe hearing loss, aided and unaided scores may not be well correlated (Franks & Jacob, 2019), but it is unknown clinically how often aided testing to demonstrate hearing aid benefit is performed, or how severe hearing loss is managed. Generally, as 4FPTA declines, aided WRS decreases (Hoppe et al., 2016; Hoppe et al., 2021; Souza et al., 2018), resulting in an increased number of poorer outcomes with hearing aids (Hoppe et al., 2021). Auditory performance with hearing aids is still variable, particularly for individuals with severe hearing loss (Souza et al., 2018). Aided speech perception in those with 4FPTA between 70 and 80 dB HL can range from 0% to 100% (Hoppe et al., 2016) and can be poorer than the “unaided” WRS under headphones by over 10% (Dorfler et al., 2020; McRackan et al., 2016). Given the variability in aided benefit with severe hearing loss, other treatment options may be tried, including assistive listening technology such as remote microphones and use of captioning devices. Visual cues in the form of captioning or lip/speech reading may be employed as additional strategies for communication when amplification is inadequate. When speech recognition is poor, cochlear implants (CI) can be considered (Birman & Sanli, 2020; Chen et al., 2017) to restore hearing (Wilson & Dorman, 2008). Since 1995, indications for cochlear implants have included patients with severe levels of hearing loss (“NIH consensus conference. Cochlear implants in adults and children,” 1995). Because of the interperson variability in auditory performance and speech recognition, it is not obvious from the clinical audiogram alone whether an individual with poor WRS would meet cochlear implant candidacy criteria. Best aided scores with use of amplification (Sydlowski et al., 2021) are not assessed during routine audiological visits, which may complicate identifying appropriate individuals who would benefit from treatments such as cochlear implants.
Severe hearing loss is prevalent among Veterans, and many with this disability receive hearing care through the VA. Those with poor WRS are less likely to benefit from traditional amplification approaches compared to those with lesser degrees of hearing loss and better preserved speech recognition abilities. This population will require alternative approaches and treatment strategies with consideration of personal preferences to support their communication needs. Future work should provide a more comprehensive depiction of how Veterans with severe hearing loss are managed within the Veteran Health Administration.
Supplemental Material
Supplemental material, sj-docx-1-tia-10.1177_23312165241273393 for Prevalence and Characteristics of Veterans with Severe Hearing Loss: A Descriptive Study by David R Friedmann, Andrew Nicholson, Colleen O'Brien-Russo, Scott Sherman and Joshua Chodosh in Trends in Hearing
Acknowledgments
The authors wish to thank Pamela E. Souza, PhD for comments on the revised manuscript.
Footnotes
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the United States government.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the U.S. Department of Veterans Affairs (grant number 5I21HX003141-05).
ORCID iD: David R Friedmann https://orcid.org/0000-0002-9136-6483
Supplemental Material: Supplemental material for this paper is available online.
References
- Atcherson S. R., Mendel L. L., Baltimore W. J., Patro C., Lee S., Pousson M., Spann M. J. (2017). The effect of conventional and transparent surgical masks on speech understanding in individuals with and without hearing loss. Journal of the American Academy of Audiology, 28(1), 058–067. 10.3766/jaaa.15151 [DOI] [PubMed] [Google Scholar]
- Beck L. B. (2010, March). A 21st century approach to VA audiology care. Association of VA Audiologists.
- Billings C. J., Dillard L. K., Hoskins Z. B., Penman T. M., Reavis K. M. (2018). A large-scale examination of veterans with normal pure-tone hearing thresholds within the department of veterans affairs. Journal of the American Academy of Audiology, 29(10), 928–935. 10.3766/jaaa.17091 [DOI] [PubMed] [Google Scholar]
- Birman C. S., Sanli H. (2020). Cochlear implant outcomes in patients with severe compared with profound hearing loss. Otology & Neurotology, 41(4), e458–e463. 10.1097/MAO.0000000000002574 [DOI] [PubMed] [Google Scholar]
- Carhart R., Porter L. S. (1971). Audiometric configuration and prediction of threshold for spondees. Journal of Speech and Hearing Research, 14(3), 486–495. 10.1044/jshr.1403.486 [DOI] [PubMed] [Google Scholar]
- Centers for Disease, C., & Prevention (2011). Severe hearing impairment among military veterans–United States, 2010. MMWR. Morbidity and Mortality Weekly Report, 60(28), 955–958. [PubMed] [Google Scholar]
- Chen S. Y., Grisel J. J., Lam A., Golub J. S. (2017). Assessing cochlear implant outcomes in older adults using HERMES: A national web-based database. Otology & Neurotology, 38(10), e405–e412. 10.1097/MAO.0000000000001575 [DOI] [PubMed] [Google Scholar]
- Cooper S. P., Alamgir H., Whitworth K. W., Gorrell N. S., Betancourt J. A., Cornell J. E., …Packer M. D. (2014). The department of defense epidemiologic and economic burden of hearing loss study. Military Medicine, 179(12), 1458–1464. 10.7205/MILMED-D-14-00100 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dillard L. K., Saunders G. H., Zobay O., Naylor G. (2020). Insights into conducting audiological research with clinical databases. American Journal of Audiology, 29(3S), 676–681. 10.1044/2020_AJA-19-00067 [DOI] [PubMed] [Google Scholar]
- Dorfler C., Hocke T., Hast A., Hoppe U. (2020). Speech recognition with hearing aids for 10 standard audiograms : English version. HNO, 68(Suppl 2), 93–99. (Sprachverstehen mit Horgeraten fur 10 Standardaudiogramme.). 10.1007/s00106-020-00843-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dubno J. R., Lee F.-S., Matthews L. J., Mills J. H. (1997). Age-related and gender-related changes in monaural speech recognition. Journal of Speech, Language, and Hearing Research, 40(2), 444–452. 10.1044/jslhr.4002.444 [DOI] [PubMed] [Google Scholar]
- Faulkner A., Rosen S., Moore B. C. (1990). Residual frequency selectivity in the profoundly hearing-impaired listener. British Journal of Audiology, 24(6), 381–392. 10.3109/03005369009076579 [DOI] [PubMed] [Google Scholar]
- Franks Z. G., Jacob A. (2019). The speech perception gap in cochlear implant patients. Cochlear Implants International, 20(4), 176–181. 10.1080/14670100.2019.1582165 [DOI] [PubMed] [Google Scholar]
- Goman A. M., Lin F. R. (2016). Prevalence of hearing loss by severity in the United States. American Journal of Public Health, 106(10), 1820–1822. 10.2105/AJPH.2016.303299 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Groenewold M. R., Tak S., Masterson E. (2011). Severe hearing impairment among military veterans–United States, 2010. Centers for Disease Control and Prevention Morbidity and Mortality Weekly Report (MMWR), 60(28), 955–958. [PubMed] [Google Scholar]
- Historical VA Hearing Aids Procurement Distribution Summaries. Retrieved June 1, 2024 from. https://department.va.gov/administrations-and-offices/acquisition-logistics-and-construction/freedom-ofinformation-act-requests/toc_Historical_VA_Hearing_Aids_Procurement_Distribution_Summaries . [Google Scholar]
- Hoppe U., Hocke T., Hast A., Iro H. (2021). Cochlear implantation in candidates with moderate-to-severe hearing loss and poor speech perception. The Laryngoscope, 131(3), E940–E945. 10.1002/lary.28771 [DOI] [PubMed] [Google Scholar]
- Hoppe U., Hocke T., Iro H. (2021). Speech Perception Across The Lifespan by Means of Artificial Intelligence. Research Square. [Google Scholar]
- Hoppe U., Hocke T., Muller A., Hast A. (2016). Speech perception and information-carrying capacity for hearing aid users of different ages. Audiology and Neurotology, 21(Suppl 1), 16–20. 10.1159/000448349 [DOI] [PubMed] [Google Scholar]
- Humes L. E., Roberts L. (1990). Speech-recognition difficulties of the hearing-impaired elderly: The contributions of audibility. Journal of Speech, Language, and Hearing Research, 33(4), 726–735. 10.1044/jshr.3304.726 [DOI] [PubMed] [Google Scholar]
- Jerger J. (1973). Audiological findings in aging. Advances in Oto-Rhino-Laryngology, 20, 115–124. Karger, Basel. 10.1159/000393092 [DOI] [PubMed] [Google Scholar]
- Lin F. R., Niparko J. K., Ferrucci L. (2011). Hearing loss prevalence in the United States. Archives of Internal Medicine, 171(20), 1851–1852. 10.1001/archinternmed.2011.506 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lin F. R., Reed N. S. (2021). The pure-tone average as a universal metric—knowing your hearing. JAMA Otolaryngology–Head & Neck Surgery, 147(3), 230–231. 10.1001/jamaoto.2020.4862 [DOI] [PubMed] [Google Scholar]
- Lyxell B., Andersson U., Borg E., Ohlsson I. S. (2003). Working-memory capacity and phonological processing in deafened adults and individuals with a severe hearing impairment. International Journal of Audiology, 42(Suppl 1), 86–89. 10.3109/14992020309074628 [DOI] [PubMed] [Google Scholar]
- McRackan T. R., Ahlstrom J. B., Clinkscales W. B., Meyer T. A., Dubno J. R. (2016). Clinical implications of word recognition differences in earphone and aided conditions. Otology & Neurotology, 37(10), 1475–1481. 10.1097/MAO.0000000000001205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mulrow C. D., Aguilar C., Endicott J. E., Tuley M. R., Velez R., Charlip W. S., …DeNino L. A. (1990). Quality-of-life changes and hearing impairment. A randomized trial. Annals of Internal Medicine, 113(3), 188–194. 10.7326/0003-4819-113-3-188 [DOI] [PubMed] [Google Scholar]
- NIH consensus conference (1995). Cochlear implants in adults and children. JAMA: The Journal of the American Medical Association, 274(24), 1955–1961. 10.1001/jama.1995.03530240065043 [DOI] [PubMed] [Google Scholar]
- Pichora-Fuller M. K., Souza P. E. (2003). Effects of aging on auditory processing of speech. International Journal of Audiology, 42(Suppl 2), 2S11–2S16. [PubMed] [Google Scholar]
- Quan H., Sundararajan V., Halfon P., Fong A., Burnand B., Luthi J. C., …Ghali W. A. (2005). Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Medical Care, 43(11), 1130–1139. 10.1097/01.mlr.0000182534.19832.83 [DOI] [PubMed] [Google Scholar]
- Radomski T. R., Zhao X., Lovelace E. Z., Sileanu F. E., Rose L., Schwartz A. L., …Thorpe C. T. (2022). Use and cost of low-value health services delivered or paid for by the veterans health administration. JAMA Internal Medicine, 182(8), 832–839. 10.1001/jamainternmed.2022.2482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Saunders G. H., Dillard L. K., Zobay O., Cannon J. B., Naylor G. (2021). Electronic health records as a platform for audiological research: Data validity, patient characteristics, and hearing-aid use persistence among 731,213 US veterans. Ear & Hearing, 42(4), 927–940. 10.1097/AUD.0000000000000980 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Souza P., Hoover E., Blackburn M., Gallun F. (2018). The characteristics of adults with severe hearing loss. Journal of the American Academy of Audiology, 29(8), 764–779. 10.3766/jaaa.17050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sydlowski S. A., King M., Petter K., Bachmann M. L. (2021). Functional assessment of hearing aid benefit: Incorporating verification and aided speech recognition testing into routine practice. Seminars in Hearing, 42(4), 365–372. 10.1055/s-0041-1739369 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Veterans Benefits Administration. (2021). VBA Annual Benefits Report Fiscal Year 2020 - Compensation. https://www.benefits.va.gov/REPORTS/abr/docs/2020_compensation.pdf
- Wang J. Z., Dhanireddy P., Prince C., Larsen M., Schimpf M., Pearman G. (2021). 2020 Survey of Veteran Enrollees’ Health Use of Health Care.
- Wells T. S., Seelig A. D., Ryan M. A., Jones J. M., Hooper T. I., Jacobson I. G., Boyko E. J. (2015). Hearing loss associated with US military combat deployment. Noise and Health, 17(74), 34–42. 10.4103/1463-1741.149574 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wiley T. L., Cruickshanks K. J., Nondahl D. M., Tweed T. S., Klein R., Klein B. E. (1998). Aging and word recognition in competing message. Journal of the American Academy of Audiology, 9(3), 191–198. [PubMed] [Google Scholar]
- Wilson B. S., Dorman M. F. (2008). Cochlear implants: Current designs and future possibilities. The Journal of Rehabilitation Research and Development, 45(5), 695–730. 10.1682/JRRD.2007.10.0173 [DOI] [PubMed] [Google Scholar]
- Wilson R. H., McArdle R. (2013). Characteristics of the audiometric 4,000 Hz notch (744,553 veterans) and the 3,000, 4,000, and 6,000 Hz notches (539,932 veterans). The Journal of Rehabilitation Research and Development, 50(1), 111–132. 10.1682/JRRD.2011.11.0225 [DOI] [PubMed] [Google Scholar]
- Yoshioka P., Thornton A. R. (1980). Predicting speech discrimination from the audiometric thresholds. Journal of Speech, Language, and Hearing Research, 23(4), 814–827. doi: 10.1044/jshr.2304.814 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material, sj-docx-1-tia-10.1177_23312165241273393 for Prevalence and Characteristics of Veterans with Severe Hearing Loss: A Descriptive Study by David R Friedmann, Andrew Nicholson, Colleen O'Brien-Russo, Scott Sherman and Joshua Chodosh in Trends in Hearing