Determinants of Hearing Aid Use Among Older Americans With Hearing Loss

Abstract

Background and Objectives

Hearing loss (HL) is common among older adults and is associated with significant psychosocial, cognitive, and physical sequelae. Hearing aids (HA) can help, but not all individuals with HL use them. This study examines how social determinants may impact HA use.

Research Design and Methods

We conducted an explanatory sequential mixed methods study involving a secondary analysis of a nationally representative data set, the Health and Retirement Study (HRS; n = 35,572). This was followed up with 1:1 qualitative interviews (n = 21) with community participants to clarify our findings. Both samples included individuals aged 55 and older with a self-reported HL, with or without HA. The main outcome measure was the proportion of participants with a self-reported HL who use HA.

Results and Discussion

Analysis of HRS data indicated that younger, nonwhite, non-Hispanic, lower income, and less-educated individuals were significantly less likely to use HA than their referent groups (all p values < .001). Area of residence (e.g., urban) were not significantly associated with HA use. Qualitative findings revealed barriers to HA included cost, stigma, vanity, and a general low priority placed on addressing HL by health care providers. Facilitators to obtaining and using HA included family/friend support, knowledge, and adequate insurance coverage for HA.

Implications

Many socioeconomic factors hinder individuals’ ability to obtain and use HA, but these obstacles appeared to be mitigated in part when insurance plans provided adequate HA coverage, or when their family/friends provided encouragement to use HA.

Hearing loss (HL), the second most common disability, affects 17% of Americans (Agrawal, Platz, & Niparko, 2008; Cruickshanks et al., 1998; Ries, 1994) and occurs predominantly in older persons: 29% of people 50–59, 45% of those 60–69, 68% of those 70–79, and 89% of those ≥80 years of age (Lin, Niparko, & Ferrucci, 2011). The adverse effects of HL are well-documented and include both reduced ability to perform independent activities of daily living and reduced quality of life (Gopinath et al., 2012). These individuals struggle with social isolation, lower income (Agrawal et al., 2008), reduced cognitive function (Wallhagen, Strawbridge, & Shema, 2008), poorer physical and psychological health (McKee, Stransky, & Reichard, 2018; Solheim, Kværner, & Falkenberg, 2011), increased risk of falls and hospitalization (Danermark & Gellerstedt, 2004; Genther, Frick, Chen, Betz, & Lin, 2013; Lin, Thorpe, Gordon-Salant, & Ferrucci, 2011), and worse patient–physician communication (Mick, Foley, & Lin, 2014). HL remains underdiagnosed and untreated. Multiple studies show that at least 75% of persons with HL do not receive treatment despite the fact that hearing aid (HA) use has been shown to reduce communication barriers, as well as disability-related effects of HL (Cacciatore et al., 1999; Metselaar et al., 2009). There remain substantial barriers to HA acquisition, most notably financial. Medicare and most health insurance plans do not provide coverage for HA, which are largely expected to be an out-of-pocket expense (Cohen-Mansfield & Taylor, 2004; Whitson & Lin, 2014). Considering the average cost of a pair of fitted HA in 2014 was between $2,200 and $7,000, this is a substantial barrier for many individuals (Whitson & Lin, 2014).

It is not as well understood how other social determinants, including area of residence and health literacy, may affect whether individuals use HA. The primary objective of this mixed methods study was twofold: (a) to examine the effect of socioeconomic determinants and area of residence, on HA use among individuals aged 55 and older with HL using a nationally representative sample; and (b) further explore any relevant findings through qualitative interviews with community participants with HL. A secondary objective looked at effect of health literacy on HA use using a subset of the HRS sample in which the data were available.

Methods

First, we conducted a secondary data analysis using data from the Health and Retirement Study (HRS). We then conducted qualitative interviews with a small group of community participants with HL, unrelated to the HRS sample, to understand more in depth both the dynamics underlying the quantitative results as well as the role of other factors, such as stigma, that HRS does not currently attempt to measure. This explanatory sequential mixed methods design allowed us to understand more fully the determinants of HA use than either qualitative or quantitative approaches alone would allow.

Quantitative Data Analysis

Procedures

We used publicly available data from the HRS, a nationally representative longitudinal study of older Americans. The design and history of the study are described in detail elsewhere (Sonnega et al., 2014); for our purposes, the HRS has two key questions which asked the participant to self-assess their hearing and whether they have HA. Respondents are first asked whether they ever wear a HA. This was used to determine the study (have HA) versus control samples (no HA) within a group of persons reporting HL. Respondents are then asked to assess their hearing as excellent, very good, good, fair, or poor; if they have reported use of a HA, they are asked to assess their hearing with a HA. Respondents who report fair or poor hearing or report ever wearing a HA were included in our sample of interest (e.g., individuals with self-assessed HL). A random subset of the HRS respondents also participated in the 2011 Health Care Mail Study, which contained health literacy data. The HRS includes detailed information on health, demographic background, and socioeconomic status (SES) of respondents.

HRS Sample

HL is more common among older adults, especially those aged 55 and older (Nash et al., 2011; National Institute on Deafness and Other Communication Disorders, 2016). Thus, we selected only subjects aged 55 and older who self-reported having difficulty hearing in the HRS. We used HRS for 1998–2012 because HRS became representative of older adults in 1998, and it continues to be representative of this population since then with refreshed samples. Our baseline sample included 35,572 individual-year observations who reported having HL (with HA = 13,248; without HA = 22,317; missing info on HA = 7).

Dependent and Explanatory Variables

Our outcome measure is a dichotomized variable which is indicative of the respondent using HA. Our explanatory variables of interest were urban–rural status of respondent’s county, SES, and health literacy. The urban–rural measure is based on the National Center of Health Statistics (NCHS) classification which includes six categories ranging from the most urban areas (large central metropolitan) to the most rural areas (noncore): (a) large central metro; (b) large fringe metro (suburb); (c) medium metro; (d) small metro; (e) micropolitan; and (f) noncore. Respondents’ demographic and socioeconomic status includes education, total income (including spouse’s), total household wealth, age, race, ethnicity, and health insurance type. Respondents’ health literacy is assessed using the Rapid Estimate of Adult Literacy in Medicine-Revised (REALM), although this measure is available only for a small subset of respondents. Insurance status is stratified by age groups: uninsured, private insurance, Medicaid, and Veterans Administration (VA) insurance for age group 55–64; and Medicare only, Medicare with private insurance, Medicare with Medicaid, and Medicare with VA insurance for age 65 and older.

Statistical Analyses

We conducted a pooled cross-sectional analysis in which the unit of observation was the person-year. To assess differences in HA use by demographic and socioeconomic characteristics, we adjusted for measures of gender, age, race/ethnicity, and survey year in the multivariable logistic regression analyses. To examine differences in HA use by the urban/rural status of residential area and health literacy, we adjusted for gender, age, race/ethnicity, education, income, wealth, and survey year. To examine differences in HA use depending on insurance status, we conducted multivariable logistic regressions that were stratified by age groups, 55–64 and 65+. We adjusted for demographic measures and socioeconomic measures sequentially in the multivariable logistic regressions. All analyses accounted for the complex designs of the surveys, including cross-sectional sample weights, sampling units, and strata.

Qualitative Data Collection and Analysis

To better interpret the findings from the quantitative data analysis, we conducted semi-structured interviews with a community sample of individuals with HL. Findings from the quantitative phase were used to build the interview protocol.

Participants and Recruitment

We recruited a purposive sample of individuals with HL to ensure that community participants were diverse and aged 55 years and older. Recruitment for the interviews was done through community flyers posted at local libraries, laundromats, University of Michigan family medicine clinics, and Craigslist online. Interested participants were screened initially by our research staff member for their eligibility (i.e., aged 55 or older, presence of HL, and ability to speak English). There were five additional screening questions that asked about educational attainment, race, ethnicity, income, and HA use. These additional screening questions helped ensure a diverse sample required to further explain the findings from the HRS analyses. Initially, a total of 23 participants were scheduled for interviews but two participants were later disqualified. The disqualification was due to participants later admitting that they had normal hearing. The final qualitative sample is based on interviews with 21 participants aged 55 years and older who took part in 1:1 semi-structured interviews between the Fall of 2016 and the Spring of 2017 at various locations around Southeast Michigan. Participants were reimbursed $30 for their time.

Procedures

All volunteering participants provided written informed consent to participate. Participants completed a short demographic survey before the interview. Based on the results of our HRS analysis, we developed a semi-structured interview guide designed to learn more about the individuals’ HL background (e.g., how diagnosed), potential barriers and facilitators on HA use, and how HL may have affected them socially and physically. The interview questions were designed to probe potential explanations for why a subject may or may not use a HA. This included potential probes on social support, personal perspectives, health care interactions, and sources of HL and HA information that were missing from the HRS analyses.

Interviews were conducted by research staff trained in qualitative data collection. Interviews were conducted in community settings, were audio-recorded, and lasted approximately 40 min. The interviewer recorded field notes during each interview, including relevant emotional responses to any particular question or interview portion.

Qualitative Data Analysis

Audio recordings were transcribed and analyzed. We used an iterative process for thematic analysis of the interview transcripts and field notes: (a) familiarized ourselves with the data; (b) generated initial codes; (c) discussed and reached consensus for codes; (d) searched for themes and patterns among participants; (e) reviewed themes; (f) defined and named themes; and (g) produced a codebook to be used with any subsequent transcript. Each transcript was coded independently by the PI and by two research assistants with the use of Dedoose software. Each transcript was compared to the emerging themes and themes were adapted to best represent the data. We created a table listing qualitative themes of the specific domains of HA use determinants. The frequency means of reported barriers and facilitators for each theme expressed by each applicable participant is displayed along with highlighted quotes. The qualitative data were used to help interpret the quantitative findings as part of an explanatory sequential mixed methods design (Guetterman, Fetters, & Creswell, 2015).

Results

In our HRS data sample of 35,572 adults aged 55 and older with self-reported HL, those using HA were more likely to be older, non-Hispanic white, educated, and wealthier (p < .001; Table 1); these effects persisted in multivariable models that include all of the demographic covariates (Table 1). Gender, area of residence, and health literacy were not significant predictors of HA use in the multivariable models (Table 1). Non-Hispanic black and those with <HS degree were the least likely to use HA (odds ratio [OR] = 0.36 and 0.43, respectively; p value < .001 for both). In supplemental analyses splitting the sample into <65 and ≥65 age groups, those with the VA provided health insurance were more likely to use HA even after controlling for demographic and socioeconomic variables (OR = 2.25; p < .001 and 1.26; p = .009, respectively for <65 and ≥65 age groups). For all other health insurance types, we found that insurance was not associated with having HA after controlling for demographic and socioeconomic covariates (Table 2). Although individuals with inadequate health literacy were less likely to use HA, this result was not statistically significant (OR = 0.82; p = .56).

Table 1.

Weighted Percent and ORs of HA Use by Sociodemographic Groups

	Sample: Age 55+ community dwelling with self-reported HL (n = 35,572)
	N	Hearing aid		Adjusted
Variable		No	Yes	OR	p Value
Gendera
Male	19,915	63.7	36.3	Ref	—
Female	15,650	63.1	37	0.85	.004
Agea
55–59	3,707	84.3	15.7	Ref	—
60–64	4,519	78.4	21.6	1.47	<.001
65–69	5,315	69.9	30.1	2.32	<.001
70–74	6,006	61.6	38.4	3.36	<.001
75–79	5,831	54.4	45.6	4.52	<.001
80–84	4,883	48.1	51.9	5.84	<.001
85+	5,304	42.7	57.4	7.18	<.001
Race/ethnicitya
Non-Hispanic, white	26,961	59.9	40.1	Ref	—
Non-Hispanic, black	4,013	81.5	18.5	0.36	<.001
Non-Hispanic, other	676	74.6	25.4	0.62	.21
Hispanic	3,898	79	21.1	0.45	<.001
Educationa
No degree (<HS)	11,417	71.4	28.7	0.43	<.001
GED/HS	17,591	62.7	37.3	0.68	<.001
At least some college	6,557	54.7	45.3	Ref	—
Wealtha
Bottom 25%	9,582	74.1	25.9	0.5	<.001
25–50%	9,374	67.6	32.4	0.61	<.001
50–75%	8,678	57.8	42.2	0.84	<.001
Top 25%	7,931	53.2	46.8	Ref	—
Health literacyb
Adequate (REALM) 7–8	231	57.5	42.5	Ref	—
Inadequate (REALM) 0–6	90	66.9	33.1	0.82	.56
Area of residenceb
1. Large central metro	8,432	64.3	35.7	Ref	—
2. Large fringe metro	6,827	62	38.1	1	.976
3. Medium metro	8,348	61	39	1.14	.131
4. Small metro	2,843	60.9	39.1	1.17	.093
5. Micropolitan	4,819	65.9	34.1	1.02	.834
6. Noncore	4,073	67.9	32.1	0.87	.32

Note: Bold connotes significance with p < .01. HA = hearing aid; HL = hearing loss; HS, high school; GED, general education diploma; OR = odds ratio; REALM = Rapid Estimate of Adult Literacy in Medicine.

^aAdjusted for gender, age, race/ethnicity, and survey year.

^bAdjusted for gender, age, race/ethnicity, education, income, wealth, and survey year.

Table 2.

Weighted Percent of Hearing Aid Use by Health Insurance and ORs of Hearing Aid for Each Insurance Group

	Sample: Age 55+ community dwelling with self-reported HL (n = 35,572)
		Hearing aid		Adjusteda
Variable	N	No (%)	Yes (%)	OR	p Value
55–64	8,228	6,438 (81.4)	1,753 (18.6)
Uninsured (ref: those with any type of insurance except VA)	1,753	1,538 (87)	215 (13)	0.77	.083
Private (ref: Medicaid or Medicare)	4,523	3,617 (80.1)	906 (20)	0.72	.082
Medicaid (ref: Private)	891	788 (78.8)	103 (21.2)	0.93	.826
VA (ref: all others expect VA)	466	312 (65.3)	154 (34.7)	2.25	<.001
65+	27,337	15,141 (55.7)	11,501 (44.4)
Medicare only (ref: Medicare + Private)	14,289	8,124 (56)	6,155 (44)	0.93	.159
Medicare + Private (ref: Medicare + Medicaid)	7,801	4,067 (51.1)	3,734 (49)	1.06	.607
Medicare+ Medicaid (ref: Medicare + Private)	2,773	1,962 (68.6)	811 (31.4)	0.95	.619
Medicare + VA (ref: all others expect VA)	1,616	778 (47.9)	838 (52.1)	1.26	.009

Note: HL = hearing loss; OR = odds ratio; VA = Veterans Administration.

^aAge, gender, race/ethnicity, education, income, wealth, and survey year are controlled.

The 21 participants interviewed for the qualitative aim of this study were mostly female, white, and had at least some college education (refer to Table 3 for participant demographics). Qualitative analysis revealed more barriers to using HA than facilitators. Commonly mentioned barriers included cost, inadequate or unclear insurance coverage for HA, vanity/ego, and a general low priority placed by health care providers to address potential HL in their patients (Table 4). Facilitators to using HA included family/friend support, knowledge, and adequate insurance coverage for HA. The following sections detail findings from the quantitative and qualitative mixed methods data sets.

Table 3.

Demographic of Interviewees

Variable	n (%)	HRS, n (%)
Age
55–64	8 (38.1)	8,228 (23.1)
65–74	8 (38)	11,319 (31.8)
75–84	5 (23.8)	10,714 (30.1)
85+	N/A	5,304 (14.9)
Gender
Male	8 (38.1)	19,915 (56.0)
Female	13 (61.9)	15,650 (44.0)
Race/ethnicity
Non-Hispanic, white	15 (71.4)	26,961 (75.8)
Non-Hispanic, black	5 (23.8)	4,013 (11.3)
Non-Hispanic, other	1 (4.8)	676 (1.9)
Hispanic	0 (0)	3,898 (11.0)
Education
HS or less	6 (28.6)	29,008 (81.6)
At least some college	15 (71.4)	6,557 (18.4)
Incomea
<$25,000	7 (35)	N/Ab
$25,000–$50,000	2 (10)	N/Ab
$50,000–$75,000	3 (15)	N/Ab
>$75,000	8 (40)	N/Ab
Hearing aid
Yes	12 (57.1)	13,248 (37.3)
No	9 (42.9)	22,317 (62.7)

Note: HRS = the Health and Retirement Study; HS, high school; N/A, not applicable.

^aOne participant declined to report income.

^bWealth quartiles were used rather than income.

Table 4.

Description of Specific Domains of HA Use Determinants

	Participant experiences (qualitative findings)
Variable	Reported barriersa	Reported facilitatorsa	Theme	Representative quote
Gender
M	1.6 (0–4)	1.1 (0–3)	Minimization of need	I’m a diabetic, so I take medications and injections and I do have to care take- take care of my diet. So, take a bunch of pills every day. Why stick one more thing on that, if it isn’t needed. (Participant 4)
F	1.8 (0–7)	1.4 (0–6)	Vanity	I stopped wearing them for a couple years, ‘cause I was in church one Sunday, and this young man said, “Oh Ms. XXX!” real loud, “You got hearing aids!” And then it embarrassed me so it just kind of blew my mind and I stopped wearing them for a couple years, like I said, and that caused my hearing to get worse, because at one time, I was only wearing one, but now I have wear both of them. (Participant 23)
Race
White	1.5 (0–5)	1.5 (0–6)	Stigma	I know that eventually I’ll get the hearing checked to the point that if I do need a hearing aid, I can get one. I’m not- I’m not against getting one, I just turned 60 so I’m just kind of like mmmm (shrugs and laughs). I’m not mentally ready for one, but I know I have a problem with hearing. (Participant 5)
			Priority of other health concerns	I just think uh when I’ve seen my doctor, it’s usually about other stuff that are, you know, um … kind of more pressing. (Participant 17)
Black	2.5 (0–7)	0.8 (0–2)	Stigma	No, it’s embarrassing. It’s like, I’m getting old. (Participant 19)
Wealth
Bottom 25%	2.7 (0–7)	0.4 (0–2)	Medical costs vs other expenses	I receive social security. I don’t get any other government subsidies. So I get no help with food. I pay for my own Medicare. I uh pay Ann Arbor rent. I don’t have a subsidy. So you know, my funds are very limited. Like after I pay all my bills and everything, I have $200 a month. I have to live off that. (Participant 20)
			Complex/other chronic diseases, medical needs	My income is such that that doesn’t leave me- by the time I get through the medical, there’s no food money, gas money or anything else money. So, you know, there’s just nothing else. ... My credit card’s maxed out from five years of other medical issues (laughs), hospital things. (Participant 14)
			Cost	I couldn’t afford $5,000 for hearing aids. Period. (Participant 15)
			Navigating health insurance plan	I’m not sure if hearing aids are covered on my insurance or not ... I’m on disability and so after like 2 years, they automatically switch you to Medicare, so I believe that, I still have the Medicaid, and I think that kicks in second, but I’m still trying to navigate. (Participant 17)
25–50%	0.7 (0–2)	1.8 (0–6)	Insurance covers HA	I have good insurance, you know, there’s money there to cover it. (Participant 5)
				I got new insurance. And that allowed me to uh get another set of hearing aids. I wasn’t really ready for them … but I didn’t have a back-up set. (Participant 7)

Note: HA = hearing aid.

^aReported as frequency mean (range) per participant based on qualitative data.

Out of Pocket Costs

In our quantitative analyses (HRS), wealth was found to be an important driver of HA ownership and use. After adjusting for controls, subjects in the lowest wealth category were half as likely to use HA when compared to the highest wealth group (Table 1). This was explored in depth in our 1:1 interviews with community respondents. Those revealed that cost was the most commonly mentioned barrier, along with inadequate or absent insurance coverage for HA. For lower income interview participants (<$25,000 annual household income), there were frequent discussions about the need to prioritize income and resources to cover essentials such as rent and food before HA.

Participant 14: My income is such that that doesn’t leave me—by the time I get through the medical, there’s no food money, gas money or anything else money. So, you know, there’s just nothing else (for me to get hearing aids). (lines 164–166)

For participants reporting <$25,000 annual household income, HA ownership facilitators were resources that are available for low-income groups or those looking for employment opportunities. For example, vocational rehabilitation provided individuals with HA.

Participant 12: [vocational rehab service] ended up paying for [hearing aids]. It was according to my income, but at the time I was not working at all, so therefore they paid for it. (lines 425–445)

Several participants commented on the lack of HA coverage with their public insurance, notably Medicaid and Medicare. Private insurance coverage for HA was critical for many of the interviewees’ ability to acquire HA regardless of their income classification. Most participants in the 1:1 interviews commented on the limitations of their insurance coverage related to HA. Medicare does not cover HA, Medicaid does in only a few states (but not in Michigan), and private plans may or may not cover HA. Participants with private coverage for HA reported that it was critical for their ability to obtain HA that they might not have been able to afford otherwise.

Participant 15 (in response what was most helpful to you in obtaining the hearing aids): Finding out that my insurance would cover ‘em, because of the expense. I know a lot of people with hearing difficulties; Medicare will not pay for hearing aids. And uh I just happened to be lucky that I retired from [xxx] and have private insurance... (lines 197–207)

Stigma of HL

Non-Hispanic blacks reported the lowest HA use of any race and ethnic groups (AOR = 0.35, p value < .001). In subsequent interviews with black participants, the stigma of HL was especially palpable, much more so than with other participants reporting other racial backgrounds. One black male added he declined hearing testing and HA fitting since he felt it would affect “his ego, identity, self-worth, [attributes] of a man” (participant 22). Even for the one black participant who was able to obtain a HA, there was a strong sense of embarrassment that prevented her from using her HA for a number of years (refer to below).

Participant 23: …I stopped wearing them for a couple of years, cause I was in church one Sunday, and this young man said, “Oh Ms. XXX!” real loud, “You got hearing aids!” And then it embarrassed me so it just kind of blew my mind and I stopped wearing them for a couple years, like I said, and that caused my hearing to get worse.

Unlike the HRS findings, qualitative analyses also suggested some differences between individuals by gender. For example, female participants described vanity as a barrier to HA use, while male participants were more likely to minimize their HL or state that they could manage without HA.

Participant 5 (female): I was only in my 50s then, so I didn’t want to do anything about it, but now I hit 60 this year, maybe it’s time I do something so I can hear. . . . But think mostly it’s just that mentality of, just not- the hearing aid means I’m really old. I’m not ready to be really old. (lines 110–111 and 129–130)

Participant 4 (male): The largest issue is communicating with my wife. I- I’m not too proud to admit that I have some hearing loss, but I don’t think I have as much as she thinks I have. That’s why I went to get the [hearing] tests last year. (lines 29–31) . . . they [audiologist] said it was mild loss and they thought it would be consistent with- they also told me I had tinnitus, which I know I have. . . They said it was consistent with aging, that uh, so I asked them, “Do I need hearing aids?” And they said, “You might benefit from them” but they didn’t say you would [definitely] benefit from them. So based on that, I didn’t get ‘em. (lines 93–98)

Health Care Providers

Primary care providers (PCP) facilitated the process of acquiring HA for several of the participants who reported higher income. Participants reported that helpful actions by their PCPs included asking or talking about the HL, including its impact on the individual’s life, setting up a referral to audiology, and/or providing additional resources such as the Hearing Loss Association of America.

Participant 15: Finding an audiologist is probably the primary thing. You know, who do you trust. Who do you, you know, wanna go see. And I had nowhere, I had no idea where to start. (lines 234–236)

Participant 12: Dr. [xxx] initially referred me over to [xxx] way back in the 80s. I’ve been her patient for a long time. A [referral] came about through our conversation [on hearing loss]. (lines 493–504)

For some participants, if the PCP did not initiate this discussion, no action was taken to address the HL, thereby affecting their ability to acquire HA.

Participant 19: (Interviewer—Have you ever talked with your primary care provider hearing concerns?)—No. It’s embarrassing . . . it’s like I’m getting old.

This contrasted with more negative PCP feedback among participants with lower income. These participants reported that PCPs not only failed to screen, but were dismissive or not helpful explaining about the importance of addressing HL. This posed a substantial barrier for subjects who did not get tested and acquire HA.

Participant 5: I think the thing the doctor could of done is at a certain age, give me information about, [hearing loss], what to do or what to look for, that type of thing . . . The only for it is that audiology test, and unless you think you have a problem with it, there’s no reason to get it. To my knowledge, um, and I’ve worked in the doctor’s office, the PCP’s office for 12 years, and I . . . I don’t know if anybody, check the hearing, but it’s rare. Nothing like the eyes or the feet or you know, other things. (lines 577–582)

Participant 22. (in response to if the subject has ever talked with the provider about hearing concerns): Yes. Interviewer: Okay, what have you talked about? P: Well if [hearing loss] gets worse, we’ll sound you out and all of that kind of stuff. We’ll deal with it then. (lines 279–284)

Participant 3: (in response if the primary care provider responded to the patient’s complaint of hearing loss) No, the primary physician had no [helpful] comment. They said, “Go see a hearing aid [specialist]” or something like that- very low key compared to what we’re talking about. (lines 309–322)

Participants valued health care providers’ recommendations and when the providers failed to give specific hearing centers and/or audiologists to see, the subjects appeared less inclined to complete the referral. There appeared to be an ongoing stigma attached to audiologists as salespeople trying to sell more HA rather than addressing the HL appropriately.

Participant 2: I think deciding who to go trust a- to do the exam and so on. I mean people who are businesses that do audiology testing and sell hearing aids, I think have um, they might be really good at it, but they might also wanna push you toward spending money. Now if I were doing [obtaining hearing aids], I would ask my doctor first for a recommendation, and then I would probably ask my friends [next]. (lines 269–274)

Audiology and hearing health professionals were mentioned some of the times as facilitators among those who obtained a HA. Multiple participants remarked on these professionals’ efforts to research community, government, and/or health insurance resources to get the HA discounted or subsidized. Participants commented that they were linked to vocational rehabilitation to obtain HA coverage (Participant 12), provided additional discounts to increase HA affordability, and that hearing health professionals even explored additional insurance coverage on behalf of the patient (Participant 7).

Significant Others

Participants indicated that their family members were important drivers of whether individuals underwent hearing testing and/or obtained HA. Participants reported that this often reflected desperation or frustration in response to the impact that HL had on their relationships and their ability to communicate with each other.

Participant 7: It was just through frustration of family members and friends saying you know let’s, let’s, we we gotta, we gotta get on top of this. They took me back and went through a battery of tests to find out why I have this hearing problem. (lines 123–125)

Participant 3: (Interviewer: So how did you end up then—so didn’t you do anything at the last appointment. How did you end up at the um hearing aid center, actually purchasing a hearing aid?) My wife bugged me enough, I decided it was easier to go get a hearing check- or discuss getting hearing aids than it was to um, argue with her. (lines 247–251)

Friends, especially peers who also had HL, were considered good resources to better understand how they can get appropriately tested and, more importantly, find out how they can get their HA covered to some extent. There were also comments in which friends steered subjects with HL away from obtaining HA.

Participant 9: I have a couple friends who um wear hearing aids and umm . . . my one friend has a couple of pairs, and like she has to- had to add ‘em to her renters insurance because she couldn’t afford to replace them if something happened. And the other guy I know, he’s constantly like trying to put more batteries. I think he needs new batteries, because he can’t hear even with them in, and um he’s always like trying to change the batteries, because he thinks the batteries are the problem . . . I don’t know if it’s covered by insurance and um I think it should be, because that probably keeps a lot of people from looking into it too because like my friend told me, hers are $5,000 each. And she told me it’s not covered insurance. (lines 81–202)

Discussion

Despite evidence demonstrating the benefits of HA use in improving psychosocial and communication abilities among individuals with HL (Cacciatore et al., 1999; Metselaar et al., 2009), our findings confirm that there remain substantial barriers to obtaining HA. The primary barriers identified in our study are financial inability, lack of insurance coverage for HA, vanity or stigma related to the HL or HA ownership, and lack of support and/or prioritizing by health care providers to address the HL.

HL represents a significant economic burden for individuals (Cassel, Penhoet, & Saunders, 2016), with the average expense of a single HA costing several thousands of dollars (Bainbridge & Wallhagen, 2014; Knudsen, Öberg, Nielsen, Naylor, & Kramer, 2010). The medical related costs associated with HL among older adults in the United States totaled about $3.1 billion in 2010 (Bainbridge & Ramachandran, 2014). When financial barriers were minimized, either through insurance coverage for HA or from other resources (e.g., vocational rehabilitation), HA use improved among our community respondents. Unfortunately, outside of VA health insurance, which provides full HA coverage, insurance was not associated with HA use in our supplemental HRS analyses. Medicaid, viewed as a stopgap for millions of Americans to ensure they have health care coverage, fails to provide HA coverage in most states (Hearing Loss Association of America, 2015). Medicare, the main health insurance for people 65 years of age and older, does not cover the cost of HA (Whitson & Lin, 2014) at all, despite recommendations by the National Academies of Medicine to do so (National Academies of Sciences, Engineering, and Medicine, 2016). Many participants in our qualitative interviews mentioned the lack of coverage as a major barrier to obtaining HA. This echoes the findings of the National Committee to Preserve Social Security and Medicare, who found that “70 percent of Americans between age 65 and 84 with HL are not using HA because Medicare does not cover them, and paying for the devices out-of-pocket is expensive” (Warren, 2017). There is a recent push for over-the-counter HA, as a strategy for reducing cost, but it is unclear to what extent this will be successful in overcoming many of the barriers listed above (Healthy Hearing, 2017). For example, an international study on HA use discovered barriers (e.g., HL stigma) remained even in countries that provide full HA coverage (Hougaard, Ruf, & Egger, 2013).

We also found that HL stigma continues to be a significant issue, especially among our black participants. Female respondents, regardless of their racial background, were frequently concerned about the potential negative effect of using HA on their appearance. One strategy that should be explored to address these barriers would be public service announcements or ad campaigns highlighting celebrities who use HA to address their HL. Despite the frequency in which HL occurs among older patients, many PCPs put lower priority in addressing HL (Zazove, McKee, et al., 2017; Zazove, Plegue, et al., 2017). Additionally, most PCPs are not comfortable with diagnosing and discussing HL as well (Zazove, McKee, et al., 2017; Zazove, Plegue, et al., 2017). Several participants brought this up during their interviews, describing the lack of engagement of their PCP regarding HL. This is largely reflective of the PCP’s lack of training on this topic, as well as the perception that this will unlikely be covered by insurance (Zazove, McKee, et al., 2017; Zazove, Plegue, et al., 2017).

A number of participants in our interviews also expressed concerns that audiologists and hearing health centers were focused on sales rather than hearing health. These negative experiences should be a clear reminder that hearing health providers should ensure that patients are given opportunities to express hearing complaints. For example, audiology centers need to be clear that HA, while helpful, do not completely compensate for HL or prevent its sequelae. Also, many consumers are not aware of their HA purchase rights to be able to test and return HA for a full refund or to switch to a different HA. The inconsistent HA coverage under insurance may be discouraging patients from even checking if HA are a covered benefit or not.

There are several steps that can be taken to improve HA uptake. First, the out-of-pocket cost for HA purchases is a significant issue that warrants action by policy makers, especially the Centers for Medicare and Medicaid Services. These policy makers should consider the potential benefits of covering HA especially in regards to their psychosocial, cognitive, and even physical long-term health. In addition, the association of HL with poorer health outcomes is another factor that potentially could be ameliorated if CMS provided coverage for HA. Another potentially helpful step would be to include training programs to inform PCPs on the importance of addressing HL and how best to do this. This could include more teaching in medical school or residencies, or the use of innovative online education materials. The quality of HA have improved dramatically over the past few years, offsetting much of the sequela of HL. Health care providers and organizations can help destigmatize HL by educating patients about the importance of addressing their HL through the use of HA. Public service announcements or ad campaigns can be tailored to target specific subgroups, including racial minorities, to help improve HA acceptance. This will also help patients who struggle with vanity and negative perceptions of agism and disability (Wallhagen, 2010).

Limitations

This study had some limitations. First, because HL in the HRS is self-reported, we have no objective measure of an individual’s degree, type, and laterality of HL. Individuals’ perceptions of HL severity may vary from individual to individual. However, self-reported HL has been shown in the literature to correlate well with objective measures of HL (Chou, Dana, Bougatsos, Fleming, & Beil, 2011; Croswell & Webber, 2013). A small subsample of participants also received pure tone hearing assessments through a survey wave called the ADAMS. The pure tone audiometry was assessed for both ears by HRS research staff (not audiologists) at the following frequencies: 500, 1,000, 2,000, 4,000 Hz at 40 decibels. Each frequency was scored as passed or not passed. There was a 64.2% agreement between self-reported HL and pure tone audiometry suggesting a HL (kappa = 0.31; p < .001).

Second, for the qualitative phase of this study we were unable to interview the same participants who had participated in the HRS. Instead, we used a community sample that mirrored the age restriction of the quantitative sample (55 years and older) and included both men and women with diverse educational attainment, income, and race. To ensure qualitative validity, we had an external review of the data by team members not involved in research design or data collection (Creswell & Miller, 2000). To ensure mixed methods validity (Creswell, 2009), we used the quantitative findings to identify the interesting and unexpected results that was subsequently explored qualitatively. Future research could expand upon our findings to integrate quantitative and qualitative data from the same sample.

Third, despite efforts to recruit a Hispanic participant with a HL in our qualitative interviews, we were unsuccessful; we also had relatively few other types of under-represented minority ethnic groups. Further studies are needed to explain why Hispanics in HRS are less likely than non-Hispanics to have HA, and to more fully understand how other ethnic minority populations perceived HL and HAs. Finally, we had no way to determine from the HRS survey how much participants used their HA.

Conclusion

In conclusion, our study identified a variety of potential barriers and facilitators regarding HA use. The most prominent were cost, lack of insurance coverage, stigma, vanity, and poor understanding and support of getting HL care by PCPs. In addition, under-represented minorities were less likely to have HA; we found a strong stigma of HA among African Americans. It appears that certain sociodemographic characteristics also predisposed respondents to HA use, including higher income, nonminority racial/ethnic status, and higher education. Further research is needed to better understand how to improve the hearing health care of persons with HL.

Funding

The work was supported by a pilot grant from the University of Michigan Geriatrics Center/Claude D. Pepper Older Americans Independence Center AG-024824/Michigan Institute for Clinical and Health Research (MICHR). The Health and Retirement Study is a longitudinal project sponsored by National Institute on Aging (Grant Number: NIA U01AG009740) and the Social Security Administration.

Conflict of Interest

None reported.