Hearing Loss and Older Adults’ Perceptions of Access to Care

Abstract

We investigated whether hard-of-hearing older adults were more likely to report difficulties and delays in accessing care and decreased satisfaction with healthcare access than those without hearing loss. The Wisconsin Longitudinal Study (2003–2006 wave, N = 6,524) surveyed respondents regarding hearing, difficulties/delays in accessing care, satisfaction with healthcare access, socio-demographics, chronic conditions, self-rated health, depression, and length of relationship with provider/site. We used multivariate regression to compare access difficulties/delays and satisfaction by respondents’ hearing status (hard-of-hearing or not). Hard-of-hearing individuals comprised 18% of the sample. Compared to those not hard-of-hearing, hard-of-hearing individuals were significantly more likely to be older, male and separated/divorced. They had a higher mean number of chronic conditions, including atherosclerotic vascular disease, diabetes and depression. After adjustment for potential confounders, hard-of-hearing individuals were more likely to report difficulties in accessing healthcare (Odds Ratio 1.85; 95% Confidence Interval 1.19–2.88). Satisfaction with healthcare access was similar in both groups. Our findings suggest healthcare access difficulties will be heightened for more of the population because of the increasing prevalence of hearing loss. The prevalence of hearing loss in this data is low and our findings from a telephone survey likely underestimate the magnitude of access difficulties experienced by hard-of-hearing older adults. Further research which incorporates accessible surveys is needed. In the meantime, clinicians should pay particular attention to assessing barriers in healthcare access for hard-of-hearing individuals. Resources should be made available to proactively address these issues for those who are hard-of-hearing and to educate providers about the specific needs of this population.

Introduction

Hearing loss is prevalent in older adults and is the sixth most common chronic condition in the United States [1]. Thirty-seven million adults in the Unites States are hard-of-hearing (have some auditory capacity) or are deaf [2]. In Wisconsin, an estimated 46% of the population over age 47 has hearing loss [3]. Individuals with other chronic conditions and physical disabilities report decreased access to care [4] and are less satisfied with their health care [5–8]. Both adults who are hard-of-hearing or deaf [9–11] as well as the physicians who treat them [12] report mutual communication difficulties in the health care setting (e.g., medication safety risks created by problems communicating and understanding a therapeutic plan). As patient-centered communication is an essential element of a satisfactory patient-physician relationship [13–15], these findings have implications for access to care. Better access to care and satisfaction with care are linked to several beneficial patient outcomes including increased adherence [16], improved receipt of preventive services [17], and higher quality of care [18, 19].

Though individuals who are deaf are known to have difficulties and delays in accessing care [20, 21], much less is known about access to health care for hard-of-hearing individuals. This group, which is increasing in size partially due to the aging of the population [22, 23], may be at heightened risk for poor access to care. Older adults, a population disproportionately affected by hearing loss, have increased morbidity from other chronic conditions. This increased morbidity compounded with communication issues in the health care setting can lead to serious safety concerns. One prior study, limited to Medicare beneficiaries, found that “hard-of-hearing” and “deaf/very hard of hearing” individuals were more dissatisfied with access to care than those with no or minor hearing difficulties [24]. However, this sample contained a number of proxy respondents, and the research analyzed as separate groups those who were “hard-of-hearing” and “deaf/very hard-of-hearing.”

Our study objective was to investigate whether older adults who are hard-of-hearing are more likely than other adults to report experiencing difficulties and delays in accessing care and decreased satisfaction with access to care.

Methods

Sample

We used data from the Wisconsin Longitudinal Study (WLS) survey, a long-term cohort study of a one-third random sample (N = 10,317) of men and women who graduated from Wisconsin high schools in the spring of 1957 and 8,778 of their randomly selected siblings [25]. In the 2003–2006 round of data collection, all surviving WLS participants were contacted via telephone. Respondents were traced and interviews were conducted and audio-recorded using computer-assisted techniques by the University of Wisconsin Survey Center. Interviews lasted approximately one hour. The response rate for this survey was 80% for graduates and 78% for siblings. Telephone interviews were followed by 54-page mail-out, mail-back surveys that took approximately 90 minutes to complete. Three mailings and one final telephone contact were made to encourage respondents to mail back their questionnaires. Among those who completed the telephone interview, a response rate of 88% for graduates and 81% for siblings was achieved.

We included participants who responded to both the telephone and mail survey and answered questions themselves about their hearing capability on the mail survey. We excluded 181 individuals who lacked insurance because this group was too small for analysis. The final sample size was 6,524. The study was approved by the Institutional Review Board at the participating university.

Variables/Measures

All variables were obtained through respondent self-report. Individuals were categorized as hard-of-hearing if they used hearing aids regularly but still reported problems hearing conversations in person or on the phone during the previous year. As only one in five people who would benefit from a hearing aid use one [26], we included as hard-of-hearing those who reported these problems and did not use a hearing aid regularly. We excluded individuals from the hard-of-hearing category who reported using hearing aids regularly and no problems with hearing conversations.

Items used to construct the two primary dependent variables, difficulties or delays in obtaining health care and satisfaction with access to care, are shown in Table 1. Items for satisfaction with access to care were from an eleven item subscale from the Group Health Association of America (GHAA) Satisfaction Survey [27].

Table 1.

Items used to construct difficulties or delays in obtaining health care and satisfaction with access to care dependent variables

Difficulties or delays in obtaining health carea

In the past 12 months, did you experience difficulty or delay in obtaining any type of health care, or not receive health care that you thought you needed due to any of the reasons listed below?

Couldn’t afford medical care

Insurance company wouldn’t approve

Cover or pay for care

Insurance required a referral, but couldn’t get one

Doctor refused to accept insurance plan

Medical care was too far away

Too expensive to get there

Couldn’t get there when doctor’s office was open

Didn’t know where to get care

Took too long to get an appointment

Couldn’t get through on telephone to make an appointment

Satisfaction with access to careb,c

Thinking about your own health care, how would you rate the following?

Convenience of doctor’s office

Hours when the doctor’s office is open

Access to specialty care if needed

Access to hospital care if needed

Access to medical care in an emergency

Access to mental health care if needed

Arrangements for making medical appointments by phone

Length of time spent waiting at the office to see the doctor

Length of time between making appointment for routine care and day of visit

Availability of medical information or advice by phone

Access to medical care whenever needed

Services available for getting prescriptions filled

^aResponse is yes/no

^bResponse on a 1–5 scale (poor, fair, good, very good, excellent)

^cItems for satisfaction with access to care were from an eleven item subscale from the Group Health Association of America (GHAA) Satisfaction Survey [27]

We adjusted analyses for potentially confounding variables. Socio-demographic information included age, gender, marital status, education, total household income, and type of health insurance. Urban or rural residence was constructed using 2004 Rural–Urban Commuting Area (RUCA) codes [28].

Other variables relating to health and physician-patient relationship included the length of relationship with a usual provider in years, the length of relationship with a usual place of care in years, and self-rated health. Self-rated health was assessed with the question, “How would you rate your health at the present time?” Responses were on a 1–5 Likert scale (very poor, poor, fair, good, excellent) and included as a continuous variable. We constructed a count of the following 22 chronic conditions: asthma, bronchitis/emphysema, serious back trouble, circulation problems, kidney/bladder problems, ulcer, allergies, multiple sclerosis, high blood pressure, diabetes, cancer, coronary heart disease/myocardial infarction, stroke, arthritis, pain and stiffness in the joints, mental illness, chronic sinusitis, fibromyalgia, high cholesterol, irritable bowel syndrome, osteoporosis and prostate problems. We also examined diabetes and a count of atherosclerotic vascular disease conditions (high blood pressure, coronary heart disease/myocardial infarction, circulation problems, stroke, high cholesterol) separately due to the association of these conditions with acquired hearing loss [29–31].

Other included variables known to be related to acquired hearing loss and lower levels of access to care included smoking status [32–34] and depressive symptoms [35–37]. Depressive symptoms were measured using the Center for Epidemiological Studies Depression Scale [38], and then dichotomized (<16, >16). A score greater or equal to sixteen is indicative of clinically significant depressive symptoms [39].

Statistical Analysis

Data were analyzed using Stata 10.0 [40]. Initial analyses included means and percentages for all variables for the sample overall. Between hearing group differences were compared by chi-square for categorical variables and one-way analysis of variance for continuous variables. Next, using logistic regression, odds ratios and 95% confidence intervals were calculated for the difficulties and delays in accessing care and linear regression was used to estimate Betas and 95% confidence intervals for satisfaction with access to care. Regression models compared individuals who were hard-of-hearing to those who were not hard-of-hearing for each dependent variable alone, and then in multivariable models that include all covariates simultaneously (age, gender, marital status, education, total household income, type of health insurance, urban or rural residence, length with a usual place, length with a usual provider, chronic conditions, smoking status, self-rated health, and depressive symptoms). Although the two continuous covariates, self-rated health and a count of chronic conditions were not normally distributed, results did not change significantly if these variables were entered in the model as categorical. We accounted for clustering of siblings within families by calculating confidence intervals and significance tests using the Stata ‘robust’ command, clustering family explicitly [41–43]. Wald tests were conducted to determine the statistical significance for groups of indicator variables. Results were considered statistically significant at a P value <0.05.

Results

Table 2 indicates the study population socio-demographic characteristics overall and by hearing capability. Eighteen percent of individuals in the sample were hard-of-hearing. The 1,203 individuals who were hard-of-hearing differed significantly from the 5,321 who were not hard-of-hearing. Those who were hard-of-hearing were older, more likely to be male, separated/divorced, and to have Medicare insurance as compared to private or other public insurance. The length of a relationship with a usual place or usual provider did not vary by hearing status (data not shown).

Table 2.

Demographics, health factors, and health care access by hearing capability

	Overall population (N = 6,524)	By hearing capability
		Not hard-of-hearing (N = 5,321)	Hard-of-hearing (N = 1,203)	P value
Mean age (SD)	64 (5)	64 (5)	65 (5)	<0.01
40–54	5	5	4
55–64	46	47	43
65+	49	48	53
Female	53	55	41	<0.001
Marital status				<0.01
Married	79	79	78
Separated/divorced	10	10	12
Widowed	7	7	7
Never married	4	4	2
Educational attainment				0.23
High school or less	52	51	52
Some college	17	17	17
College	15	15	16
Post-graduate	16	16	14
Total household income ($)				0.34
Less than $30,000	17	17	17
$30,000–$44,999	15	15	15
$45,000–59,999	13	13	13
$60,000–$74,999	11	11	11
Greater than $75,000	30	31	29
Missing	13	13	15
Rural residence	33	32	35	0.28
Health insurance				0.03
Private	47	48	44
Medicare + other private	40	39	42
Medicare	12	11	13
Other public	2	2	1
Ever a regular smoker	55	54	57	0.08
Number of chronic conditions mean (SD)a	3.6 (2.5)	3.5 (2.5)	4.2 (2.7)	<0.001
Atherosclerotic vascular disease	49	47	55	<0.001
Diabetes	12	12	15	<0.01
Self-rated health mean (SD)b	4.0 (0.69)	4.0 (0.68)	3.9 (0.71)	<0.001
High depressive symptomatologyc	35	33	46	<0.001
Difficulties/delays in health care access	9	8	13	<0.001
Satisfaction with access to care mean (SD)d	3.6 (0.7)	3.7 (0.7)	3.6 (0.7)	<0.001

Values represent percents unless specified otherwise

^aThe following 22 chronic conditions were measured in this count: asthma, bronchitis/emphysema, serious back trouble, circulation problems, kidney/bladder problems, ulcer, allergies, multiple sclerosis, high blood pressure, diabetes, cancer, coronary heart disease/myocardial infarction, stroke, arthritis, pain and stiffness in the joints, mental illness, chronic sinusitis, fibromyalgia, high cholesterol, irritable bowel syndrome, osteoporosis and prostate problems

^bSelf-rated health was assessed with the question, “How would you rate your health at the present time?” Responses were on a 1–5 Likert scale (very poor, poor, fair, good, excellent)

^cDepressive symptoms were measured using the Center for Epidemiological Studies Depression Scale [38], and a score greater or equal to sixteen as indicative of clinically significant depressive symptoms [39]

^dSummary measure of 11 items where 1 = poor and 5 = excellent

Table 2 also shows the difference in utilization and health characteristics by hearing capability, and overall. Individuals who were hard-of-hearing differed from those not hard-of-hearing by having a higher average number of chronic conditions. They were proportionally more likely to have diabetes mellitus, atherosclerotic vascular disease, clinically significant depressive symptoms, and had slightly lower self-rated health. Thirteen percent of those hard-of-hearing reported experiencing difficulties and delays in healthcare access in the past 12 months as compared to 8% of those not hard-of-hearing (P < 0.01). Satisfaction with access to care was significantly lower for those hard-of-hearing, compared to those without hearing difficulty (P < 0.01).

As shown in Table 3, after adjustment for potential confounders, hard-of-hearing individuals still were more likely to report difficulties and delays in accessing healthcare in the past 12 months. Other significant predictors of reporting difficulties and delays in accessing care were having a bachelor’s degree as compared to a high school education or less, having an increased number of chronic conditions, and having significant depressive symptoms.

Table 3.

Adjusted odds ratios and Betas with 95% confidence intervals for difficulties and delays in care and satisfaction with access to care

	Experienced difficulties/delays in care in the past 12 months		Satisfaction with access to care
	OR	95% CI	Beta	95% CI
Hard-of-hearing	1.85	(1.19, 2.88)	−0.06	(−0.15, 0.02)
65+ years old	0.99	(0.57, 1.71)	−0.07	(−0.17, 0.03)
Female	1.23	(0.80, 1.89)	0.10	(0.02, 0.17)
Marital status
Separated/divorced	1.25	(0.62, 2.52)	−0.08a	(−0.20, 0.05)
Widowed	1.52	(0.77, 3.02)	0.00	(−0.14, 0.13)
Never married	1.94	(0.85, 4.45)	−0.03	(−0.20, 0.15)
Educational attainment
Some college	1.64	(0.95, 2.83)	0.04	(−0.06, 0.15)
College	2.38	(1.34, 4.26)	0.11	(0.00, 0.22)
Post-graduate	1.39	(0.72, 2.69)	0.12	(0.00, 0.23)
Total household income ($)
$30,000–$44,999	0.98	(0.52, 1.84)	−0.05	(−0.16, 0.05)
$45,000–59,999	1.18	(0.62, 2.24)	0.03	(−0.08, 0.14)
$60,000–$74,999	0.96	(0.45, 2.04)	0.10	(−0.03, 0.23)
Greater than $75,000	1.02	(0.53, 1.95)	0.07	(−0.04, 0.17)
Missing	0.80	(0.24, 2.65)	0.04	(−0.14, 0.22)
Residence
Rural resident	0.67	(0.42, 1.05)	0.01	(−0.06, 0.08)
Health insurance
Medicare + other private	1.42	(0.79, 2.55)	0.04	(−0.07, 0.14)
Medicare	1.14	(0.48, 2.68)	0.22	(0.08, 0.36)
Other public	3.99a	(1.37, 11.6)	0.11	(−0.17, 0.39)
Ever a regular smoker	1.27	(0.84, 1.92)	0.06	(−0.01, 0.13)
Number of chronic conditionsb	1.11	(1.02, 1.22)	0.01	(−0.01, 0.02)
Atherosclerotic vascular diseasec	0.80	(0.52, 1.24)	0.03	(−0.05, 0.10)
Diabetesd	0.70	(0.40, 1.22)	0.01	(−0.09, 0.10)
Self-rated healthe	0.87	(0.62, 1.24)	0.12	(0.06, 0.17)
High depressive symptomatologyf	1.63	(1.07, 2.49)	−0.16	(−0.23, −0.08)

Bold values indicate statistical significance at P < 0.05

This model included the following covariates: age, gender, marital status, education, total household income, type of health insurance, urban or rural residence, length with a usual place, length with a usual provider, chronic conditions, smoking status, self-rated health, and depressive symptoms. Referent groups were: not hard-of-hearing, less than 65 years old, male, married, high school or less, total household income under $30,000, urban resident, private insurance, never a smoker, no chronic conditions, no atherosclerotic vascular disease, no diabetes, very poor self-rated health, and no clinically significant depressive symptoms

^aResults were not significant when tested as a group of indicator variables

^bThe following 22 chronic conditions were measured in this count: asthma, bronchitis/emphysema, serious back trouble, circulation problems, kidney/bladder problems, ulcer, allergies, multiple sclerosis, high blood pressure, diabetes, cancer, coronary heart disease/myocardial infarction, stroke, arthritis, pain and stiffness in the joints, mental illness, chronic sinusitis, fibromyalgia, high cholesterol, irritable bowel syndrome, osteoporosis and prostate problems

^cA separate model that did not include the number of chronic conditions gave the following results: atherosclerotic vascular disease OR 0.94 (0.45–1.36) and Beta 0.03 (−0.03 to 0.10)

^dA separate model that did not include the number of chronic conditions gave the following results: diabetes OR 0.78 (0.45–1.36) and Beta 0.01 (−0.08 to 0.11)

^eSelf-rated health was assessed with the question, “How would you rate your health at the present time?” Responses were on a 1–5 Likert scale (very poor, poor, fair, good, excellent)

^fDepressive symptoms were measured using the Center for Epidemiological Studies Depression Scale [38], and a score greater or equal to sixteen as indicative of clinically significant depressive symptoms [39]

Also shown in Table 3, after adjustment, satisfaction with access to care was similar in those who were and were not hard-of-hearing. Significant predictors of higher satisfaction with access to care were female gender, Medicare insurance, and higher self-rated health. Those who had significant depressive symptoms reported significantly lower satisfaction with access to care as compared to those without significant depressive symptoms.

Discussion

We found individuals who were hard-of-hearing as compared to those not hard-of-hearing to be significantly more likely to report experiencing difficulties and delays in accessing care. The disparity in access persisted even after controlling for several variables known to influence health care access and satisfaction. However, the two groups did not differ on satisfaction with access to care.

Our findings differ from prior studies that found lower satisfaction with access to care in populations with disabilities [24, 44–46]. However, our study design allowed us to include variables not included in these prior studies that are known to have an important effect on satisfaction with access to care. In particular, we adjusted for depression, which has a significant negative effect on satisfaction with access to care [47]. Our unadjusted results did reveal a statistically significant difference in satisfaction with access to care between groups (Table 2). Alternatively, differences in sample populations also may account for our varying results. In particular, nearly one-third of the responses to the Medicare Care Beneficiary Survey were completed by proxies, allowing for representation from those with more profound disabilities such as severe hearing loss [24]. The focus in our sample was on those hard-of-hearing, but able to complete their own telephone surveys. Alternatively, our sample was predominantly well educated and white. Therefore, some of our differences may be explained by the different socio-demographic factors of our sample as compared to samples drawn from nationally representative surveys.

Our findings of similar satisfaction with access to care despite significant differences in perceived difficulties and delays in accessing care adds to the literature suggesting that detailed questions about barriers to accessing care may be more helpful than a summary satisfaction measure. Discrepancies between overall high satisfaction scores and reported problems with care have been reported after hospitalization [48, 49] leading to a call for specificity in patient feedback when identifying problems for intervention [50].

Our findings should be considered in light of several limitations. Our sample represents individuals who were attending Wisconsin high schools in the 1950s and therefore is limited in geographical and racial/ethnic diversity. Our classification of individuals as hard-of-hearing is based on self-report and thus may be subject to misclassification bias. However, the accuracy of self-reported hearing loss has been validated in older adult populations [51]. Next, the proportion of individuals with hearing loss in our sample was substantially lower than that estimated in an age matched older adult population. This discrepancy is partially explained by the relatively high education bias and income in our sample, as hearing loss is inversely associated with these factors [3]. Also, by using data from respondents to a combined telephone and mail survey, our findings excluded hard-of-hearing individuals who could not use a telephone because of their hearing loss, did not feel comfortable answering questions on the telephone [52, 53], or did not own a phone [54]. By excluding these individuals it is likely that analyses of the Wisconsin Longitudinal Study data underestimate actual disparities in access to care. Further research in this area incorporating methods other than telephone surveys is needed.

In conclusion, our findings suggest that adults who are hard-of-hearing are more likely to experience difficulties and delays in accessing health care as compared to those who are not hard-of-hearing. Further investigation is needed into why hard-of-hearing individuals report more difficulties and delays in accessing care. This research could examine the age of onset of hearing difficulty in relation to differences in health care behavior, and incorporate survey methods other than telephone surveys. Our findings are concerning for multiple reasons. First, with the increasing prevalence of hearing loss [23], access to care issues will be heightened for more of the population. Furthermore, as there is significant co-morbidity associated with hearing-loss, poor access [55, 56] compounded by known communication difficulties in the physician-patient relationship [9–12] poses serious health risks. Clinicians should pay particular attention to assessing barriers in access to care for hard-of-hearing individuals. Resources should be made available to proactively address the access to care issues for those who are hard-of-hearing and to educate providers about the specific needs of this population.