Abstract
Several studies have recently illuminated the relationship between hearing loss and hospitalizations, but little is understood as to why hearing loss is associated with greater risk for hospitalizations. This study examines the role of patient–provider communication as the mechanism by which those with hearing loss are more likely to be hospitalized, using the self-reported data from 12,654 Medicare beneficiaries from the 2016 Medicare Current Beneficiary Survey (MCBS) Cost and Use File. Multivariable logistic regression was used to model the odds of any hospitalization in the past year and negative binomial regression to model the incident rate ratio of hospitalization based on number of hospitalizations in the past year. It was found that Medicare beneficiaries who experience impaired communication with their physician due to trouble hearing have greater odds of hospitalization and a higher rate of all hospitalization over a 1-year period. Understanding the complex relationship between hearing loss, patient–provider communication, and hospitalizations may provide health care professionals with a better rational to address not only hearing loss but also impaired communication in the care of an individual.
Keywords: hearing loss, patient–provider communication, hospitalizations, Medicare beneficiaries
Hearing loss is a prevalent, yet often overlooked public health concern. Approximately 38 million 1 Americans have hearing loss and prevalence increases with age such that it affects two-thirds of adults older than 70 years. 2 Given the aging demographics of the United States, the number of adults with hearing loss is expected to nearly double by the year 2060. 3 Hearing loss is associated with declines in cognitive, 4 social, 5 and physical functioning. 6 Recent research suggests hearing loss is also associated with health care spending and utilization. Hearing aid use may modify the association of hearing loss and negative health outcomes; however, there is a paucity of randomized control trials to assess this hypothesis. 7
Optimal patient–provider communication is a cornerstone of patient-centered care and is associated with improved health care outcomes. 8 Conversely, poor communication is associated with poor treatment understanding, feelings of being uninvolved in the decision-making process of one's care, and reduced satisfaction with care. 9 10 11 12 13 Hearing loss may be associated with health care spending and utilization via its negative impact on communication between patients and providers. Recent research suggests adults with hearing loss report difficulty understanding health professionals, especially when unfamiliar language (i.e., health jargon) is used during the appointment. 14 Moreover, providers report hearing loss contributes to poor patient–provider rapport. 9
Several recent articles have examined hearing loss and health care spending, utilization, and satisfaction. In a nationally representative sample of older adults in the United States, those with hearing loss were more likely to experience a hospitalization. 15 In an analysis of a large administrative claims dataset, researchers found that individuals with hearing loss experienced 46% higher total health care costs, a higher rate of hospitalization, and 1.44 times the risk to experience a 30-day readmission relative to those without hearing loss over a 10-year period. 16 Moreover, previous work suggests older adults with hearing loss have higher odds of reporting decreased satisfaction with quality of care. 17 An analysis of Medicare beneficiaries found that hearing aid use was associated with increased overall health care spending (i.e., Medicare insurance, supplemental insurance, and out-of-pocket) but decreased Medicare-specific spending by $71, on average, in a 1-year period. In a study of Medicare beneficiaries with hearing aids, researchers found those who reported using hearing care services in the past year spent, on average, $2,513 less than those who did not use hearing care services. 18
While some previous articles have suggested hearing loss is associated with increased hospitalization, there is little work examining self-report patient–provider communication and the influence of hearing aid use on this association. The aim of this article is to characterize the association between hospitalization and hearing loss and whether hearing aid use modifies this association in a nationally representative Medicare Current Beneficiary Survey (MCBS). Moreover, we will describe the association between hospitalization and self-report trouble communicating with medical professionals due to hearing loss. We hypothesize that hearing loss, poor communication due to hearing loss, and untreated hearing loss are all associated with increased odds of any hospitalization and an increased rate of hospitalization. However, Medicare beneficiaries without hearing loss, those who do not report hearing interferes with communication, and those who use hearing aids, respectively, are less likely to experience an increased odds of any hospitalization and an increased rate of hospitalization.
Methods
Data Source
The MCBS 19 is a panel survey covering a nationally representative sample of the Medicare population enrolled in either Part A or Part B of the Medicare program in the United States and Puerto Rico. It is executed and provided by the Centers for Medicare and Medicaid Services and includes individuals younger than 65 years with a qualifying disability enrolled in Medicare in addition to those who qualify for Medicare by age (65 and older). Interviews are completed in person with a computer-assisted interviewing program. The beneficiary or proxy in their residence or neutral interview location completes the interview. The MCBS collects information pertaining to health care use, access, and expenditures, as well as information about the beneficiary. The 2016 MCBS Cost and Use File includes 14,778 participants with varying levels of completeness of data. This study includes 12,654 participants after excluding those with missing covariates. Because the data are de-identified and complies with the Health Insurance Portability and Accountability Act, institutional review board approval was not required.
Outcome Variables
The primary outcome of this study was hospitalization as measured by number of hospitalizations in the previous year. Beneficiaries could report up to 10 hospitalizations. Hospitalizations were modeled as any hospitalization in the previous year (binary, yes or no variable) and as a continuous variable on a scale of 0 to 10 or more hospitalizations.
Exposure
To assess hearing status, Medicare beneficiaries were asked if they used a hearing aid (yes or no) and which statement best described their hearing, with hearing aids if applicable (no trouble hearing, a little trouble hearing, and a lot of trouble hearing). In follow-up, those who reported any trouble hearing were asked how much trouble they have communicating with their doctor or other health professional because of difficulty hearing (no trouble hearing, a little trouble hearing, and a lot of trouble hearing).
Given that trouble hearing includes with a hearing aid, if appropriate, we considered the exposure variable in this study as “functional hearing loss” that represents everyday conditions (i.e., with aids if appropriate). Based on participant response to the aforementioned questions, we categorized participants into one of the three categories of functional hearing loss: no hearing trouble, a little trouble hearing, and a lot of trouble hearing. Next, those with trouble hearing were categorized as having no trouble communicating, a little trouble communicating, and a lot of trouble communicating with health care providers due to hearing loss. To assess the impact of hearing aid use, we categorized participants as using a hearing aid, reporting a little trouble hearing and not using a hearing aid, and reporting a lot of trouble hearing and not using a hearing aid.
Covariates
The Andersen–Aday 20 behavioral model for health services was used to identify predisposing, enabling, and need factors. Age (modeled as a continuous variable), self-report sex (male or female), and self-report race (non-Hispanic white, non-Hispanic black, Hispanic, or other) were identified as predisposing factors. Enabling factors included income-to-poverty ratio (<100%, 101–200%, and >200%) and educational attainment (< 9th grade, high school graduate, vocational/some college/associates degree, or bachelors or higher). Need factors included chronic comorbidities count (categorized as 0, 1–2, 3–5, and >6 of the following conditions: cancer, chronic obstructive pulmonary disease/asthma, chronic heart disease, serious mental illness, acute myocardial infarction, hypertension, diabetes, stroke, depression, arthritis, dementia, or Alzheimer's disease; Table 1 ).
Table 1. Demographic and socioeconomic characteristics of Medicare beneficiaries by self-report functional hearing loss a b .
| Functional hearing status | ||||
|---|---|---|---|---|
| Variable | Sample | No trouble | A little trouble | A lot of trouble |
| Unweighted sample | 12,654 | 6,710 | 5,074 | 870 |
| Weighted sample | 57,877,758 | 55.54% | 38.72% | 5.75% |
| At least one hospitalization | 8.91% | 8.12% | 9.35% | 13.61% |
| Mean hospitalizations (95% CI) | 0.13 (0.12–0.14) | 0.11(0.10–0.13) | 0.14 (0.12–0.15) | 0.22 (0.15–0.29) |
| Mean age c (95% CI) | 70.75 (70.68–70.83) | 69.30 (69.12–69.48) | 72.26 (72.01–72.50) | 74.67 (73.71–75.63) |
| Female | 54.23% | 59.39% | 48.54% | 42.74% |
| Race | ||||
| Non-Hispanic white | 76.42% | 71.72% | 82.39% | 81.72% |
| Non-Hispanic black | 10.31% | 12.61% | 7.56% | 6.69% |
| Hispanic | 7.89% | 9.77% | 5.37% | 6.67% |
| Other | 5.38% | 5.91% | 4.68% | 4.92% |
| Educational attainment | ||||
| Less than high school | 16.60% | 16.95% | 14.70% | 25.95% |
| High school graduate | 26.87% | 26.79% | 26.56% | 29.81% |
| Vocational, some college, associates | 29.82% | 29.43% | 30.63% | 28.04% |
| Bachelors or higher degree | 26.71% | 26.83% | 28.10% | 16.20% |
| Income to poverty ratio | ||||
| Less than 100% | 15.60% | 17.39% | 12.83% | 16.92% |
| 101–200% | 25.54% | 25.38% | 24.43% | 34.53% |
| Greater than 200% | 58.87% | 57.23% | 62.75% | 48.55% |
| Chronic comorbidities count d | ||||
| 0 | 9.51% | 11.33% | 7.72% | 4.03% |
| 1–2 | 40.82% | 44.57% | 37.61% | 26.15% |
| 3–5 | 39.68% | 35.78% | 44.15% | 47.25% |
| >6 | 9.99% | 8.33% | 10.51% | 22.57% |
Data derived from the 2016 Medicare Current Beneficiary Survey Cost and Use File.
Survey weights applied according to Medicare Current Beneficiary Survey; percentages are based on weighted survey sample.
Age is in years.
Chronic comorbidities include cancer, chronic obstructive pulmonary disease/asthma, chronic heart disease, serious mental illness, acute myocardial infarction, hypertension, diabetes, stroke, depression, arthritis, and dementia or Alzheimer's disease.
Statistical Methods
Descriptive and univariate chi-square analyses were used to explore correlations and identify trends between functional hearing status and covariates, and outcome variables. Multivariable logistic regression models were used to describe the association between exposure and any hospitalization, while negative binomial regression was used to model the association between exposure variables and hospitalization as a continuous measure. Notably, negative binomial regression was used to account for excess zero counts (i.e., most participants did not experience a hospitalization). The β-coefficients (log-odds) were converted into odds ratios [ORs] for ease of interpretation of logistic models, while incident rate ratios were reported for negative binomial models. Significance testing for all analyses was two sided with a type I error of 0.05. Individuals with missing data were excluded from analyses. Survey weighting was applied to the data to account for MCBS oversampling of subpopulations and cluster sampling design. The statistical software used was Stata 15 (StataCorp, College Station, TX).
Results
Characteristics of Study Sample
After excluding those who had missing hearing data, this study included 12,654 Medicare beneficiaries in the unweighted sample and 57,877,758 in the weighted sample. The characteristics of the weighted sample are displayed in Table 1 based on self-reported hearing status. Among Medicare beneficiaries, 56% reported no trouble hearing, 39% reported a little trouble hearing, and 6% reported a lot of trouble hearing. A greater proportion of those who reported a lot of trouble hearing (57%) were males, whereas a greater proportion of those who reported no trouble hearing (59%) or a little trouble hearing (49%) were females. A greater proportion of non-Hispanic whites reported a little trouble hearing (82%) and a lot of trouble hearing (82%) compared with those who reported no trouble hearing (72%). On the other hand, a greater proportion of non-Hispanic blacks and Latino/Hispanic Medicare beneficiaries reported no trouble hearing. Furthermore, Medicare beneficiaries who reported a little trouble hearing (mean age = 72.26 years) or a lot of trouble hearing (mean age = 74.67 years) were, on average, older than those with no trouble hearing (mean age = 69.30 years). Of those who reported a lot of trouble hearing, 26% had less than high school education; 30% had high school education or equivalent; 28% attained vocational, some college, or associates degree education; and 16% attained a bachelors or higher degree. This differs from those who reported no trouble hearing or a little trouble hearing, as the greatest proportion of those individuals, 56 and 59%, respectively, had education more than high school. Also, Medicare beneficiaries who reported no trouble hearing had fewer chronic conditions relative to those with a little and a lot of trouble hearing.
Model 1: Self-Reported Functional Hearing Status
Table 2 displays the multivariable logistic regression for the odds of any hospitalization over the last year and incident rate ratio of hospitalization. Medicare beneficiaries who reported a little trouble hearing had 1.04 times the odds (OR: 1.04, 95% confidence interval [CI]: 0.89–1.23) of experiencing a hospitalization compared with those with no trouble hearing. For those who reported a lot of trouble hearing, these individuals had 1.30 times the odds (OR: 1.30, 95% CI: 1.00–1.69) of experiencing a hospitalization in the past year compared with those with no trouble hearing. Moreover, compared with Medicare beneficiaries with no trouble hearing, a little trouble hearing and a lot of trouble hearing were associated with a 5% (incident rate ratio [IRR]: 1.05, 95% CI: 0.89–1.25) and 43% (IRR: 1.43, 95% CI: 0.97–2.11) higher rate of hospitalization, although these differences were not found to be statistically significant.
Table 2. Association of self-reported functional hearing status and hospitalization among Medicare beneficiaries a b .
| Any hospitalization c | Number of hospitalizations d | |||
|---|---|---|---|---|
| Variable | Odds ratio (95% CI) | Standard error | Incident rate ratio (95% CI) | Standard error |
| Self-reported functional hearing status | ||||
| No trouble | Ref. | Ref. | ||
| A little trouble | 1.04 (0.89–1.23) | 0.09 | 1.05 (0.89–1.25) | 0.09 |
| A lot of trouble | 1.30 (1.00–1.69) * | 0.17 | 1.43 (0.97–2.11) | 0.28 |
| Age (y) | 1.01 (1.01–1.02) ** | 0.00 | 1.01 (1.00–1.01) ** | 0.00 |
| Female | 0.97 (0.83–1.13) | 0.08 | 0.96 (0.82–1.13) | 0.08 |
| Race/Ethnicity | ||||
| Non-Hispanic white | Ref. | Ref. | ||
| Non-Hispanic black | 1.15 (0.92–1.45) | 0.13 | 1.28 (1.00–1.64) * | 0.16 |
| Hispanic | 0.92 (0.64–1.30) | 0.16 | 1.00 (0.71–1.41) | 0.17 |
| Other | 0.87 (0.61–1.24) | 0.16 | 0.80 (0.56–1.15) | 0.15 |
| Education | ||||
| Less than high school | Ref. | Ref. | ||
| High school graduate | 0.99 (0.80–1.23) | 0.11 | 1.01 (0.81–1.25) | 0.11 |
| Vocational, some college, associates | 1.00 (0.78–1.29) | 0.13 | 1.02 (0.81–1.28) | 0.12 |
| Bachelors or higher degree | 0.78 (0.59–1.02) | 0.11 | 0.81 (0.61–1.07) | 0.11 |
| Income to poverty ratio | ||||
| Less than 100% | Ref. | Ref. | ||
| 101–200% | 0.90 (0.73–1.12) | 0.10 | 0.90 (0.70–1.16) | 0.12 |
| Greater than 200% | 0.81 (0.65–1.00) | 0.09 | 0.75 (0.60–0.96) * | 0.09 |
| Chronic comorbidities count e | ||||
| 0 | Ref. | Ref. | ||
| 1–2 | 1.60 (1.04–2.48) * | 0.35 | 1.93 (1.26–2.96) * | 0.41 |
| 3–5 | 3.13 (2.06–4.78) ** | 0.67 | 3.96 (2.63–5.95) ** | 0.81 |
| ≥ 6 | 3.95 (2.46–6.35) ** | 0.94 | 5.51 (3.48–8.71) ** | 1.27 |
p < 0.05.
p < 0.001.
Data derived from the 2016 Medicare Current Beneficiary Survey Cost and Use File.
Survey weights applied according to Medicare Current Beneficiary Survey; percentages are based on weighted survey sample.
Logistic regression model.
Negative binominal regression model.
Chronic comorbidities include cancer, chronic obstructive pulmonary disease/asthma, chronic heart disease, serious mental illness, acute myocardial infarction, hypertension, diabetes, stroke, depression, arthritis, and dementia or Alzheimer's disease.
Among sociodemographic characteristics, sex and level of education were not associated with any hospitalization or number of hospitalizations in the past year. However, non-Hispanic black Americans have a higher rate (IRR: 1.28, 95% CI: 1.00–1.64) of hospitalization in comparison to non-Hispanic white Americans. Each year of age was significantly associated with greater odds of any hospitalizations in the past year (OR: 1.01, 95% CI: 1.01–1.02) and significantly higher rate of hospitalizations in the past year (IRR: 1.01, 95% CI: 1.00–1.01). Lastly, the numbers of chronic comorbidities were significantly associated with increased odds of any hospitalization and rate of hospitalization in a dose–response pattern.
Model 2: Patient–Physician Communication
In the weighted sample, 13.6% of Medicare beneficiaries with self-reported functional hearing loss reported a little trouble and 2.94% reported a lot of trouble communicating with their health care provider due to hearing trouble, respectively. Table 3 displays a multivariable logistic regression for odds of any hospitalization and negative binomial regression rate of hospitalization by self-report difficulty communicating with health care providers due to hearing, among Medicare beneficiaries with functional hearing loss. Those who reported a little trouble communicating with their physician due to hearing had 28% greater odds (OR: 1.28, 95% CI: 1.01–1.63) of having any hospitalization in the past year and 36% higher rate of hospitalization (IRR: 1.36, 95% CI: 1.03–1.80) compared with those without difficulty communicating with their provider. Furthermore, those with a lot of trouble communicating with their health care providers had higher, albeit not significant, odds of any hospitalization (OR: 1.55, 95% CI: 0.97–2.49) and significantly higher rate of hospitalization (IRR: 2.40, 95% CI: 1.17–4.93) relative to those without trouble communicating with health care providers. Among socioeconomic variables, only higher income was associated with fewer hospitalizations. Results for chronic comorbidities were consistent with the trends seen in Table 2 .
Table 3. Association of self-report trouble communicating with health care providers due to hearing trouble and hospitalization among Medicare beneficiaries with functional hearing loss a b .
| Any hospitalization c | Number of hospitalizations d | |||
|---|---|---|---|---|
| Variable | Odds ratio (95% CI) | Standard error | Incident rate ratio (95% CI) | Standard error |
| Self-reported trouble communicating with medical provider | ||||
| No trouble | Ref. | Ref. | ||
| A little trouble | 1.28 (1.01–1.63) * | 0.16 | 1.36 (1.03–1.80) * | 0.19 |
| A lot of trouble | 1.55 (0.97–2.49) | 0.37 | 2.40 (1.17–4.93) * | 0.87 |
| Age (y) | 1.02 (1.01–1.03) * | 0.01 | 1.02 (1.00–1.03) * | 0.01 |
| Female | 1.01 (0.81–1.26) | 0.11 | 0.94 (0.74–1.19) | 0.11 |
| Race/Ethnicity | ||||
| Non-Hispanic white | Ref. | Ref. | ||
| Non-Hispanic black | 1.12 (0.78–1.61) | 0.20 | 0.96 (0.69–1.33) | 0.16 |
| Hispanic | 1.09 (0.71–1.65) | 0.23 | 1.16 (0.71–1.88) | 0.28 |
| Other | 1.19 (0.67–2.10) | 0.34 | 1.01 (0.56–1.83) | 0.30 |
| Education | ||||
| Less than high school | Ref. | Ref. | ||
| High school graduate | 1.06 (0.79–1.41) | 0.15 | 1.07 (0.79–1.46) | 0.17 |
| Vocational, some college, associates | 1.02 (0.74–1.41) | 0.16 | 1.16 (0.87–1.53) | 0.16 |
| Bachelors or higher degree | 0.68 (0.46–0.99) | 0.13 | 0.80 (0.52–1.19) | 0.16 |
| Income to poverty ratio | ||||
| Less than 100% | Ref. | Ref. | ||
| 101–200% | 0.95 (0.68–1.34) | 0.17 | 0.82 (0.52–1.28) | 0.18 |
| Greater than 200% | 0.85 (0.61–1.20) | 0.15 | 0.63 (0.40–0.99) * | 0.14 |
| Chronic comorbidities count e | ||||
| 0 | Ref. | Ref. | ||
| 1–2 | 3.99 (2.13–7.46) ** | 1.26 | 4.80 (2.55–9.02) ** | 1.53 |
| 3–5 | 7.37 (3.84–14.11) ** | 2.41 | 9.41 (4.98–17.75) ** | 3.01 |
| ≥ 6 | 7.97 (4.05–15.68) ** | 2.72 | 8.82 (4.53–17.16) ** | 2.96 |
p < 0.05.
p < 0.001.
Data derived from the 2016 Medicare Current Beneficiary Survey Cost and Use File.
Survey weights applied according to Medicare Current Beneficiary Survey; percentages are based on weighted survey sample.
Logistic regression model.
Negative binominal regression model.
Chronic comorbidities include cancer, chronic obstructive pulmonary disease/asthma, chronic heart disease, serious mental illness, acute myocardial infarction, hypertension, diabetes, stroke, depression, arthritis, and dementia or Alzheimer's disease.
Model 3: Hearing Aid Use
Among Medicare beneficiaries who reported functional hearing loss or hearing aid use (i.e., no trouble hearing when using a hearing aid), 25.11% reported hearing aid use, 66.26% reported a little trouble hearing and no hearing aid use, and 8.64% reported a lot of trouble hearing and no hearing aid use. Table 4 displays a multivariable logistic regression for odds of any hospitalization and negative binomial regression for rate of hospitalization among those who reported hearing trouble with no hearing aid use referenced against those who report hearing aid use. Medicare beneficiaries with a little trouble hearing without hearing aid use had similar odds to those who used hearing aids. However, those who reported a lot of trouble hearing with no hearing aid use had 59% greater odds (OR: 1.59, 95% CI: 1.18–2.16) of experiencing any hospitalization in the past year and 57% higher rate of hospitalization (IRR: 1.57, 95% CI: 0.97–2.53) relative to those who were hearing aid users. Results of other variables included in this model were consistent with previous models.
Table 4. Association of hearing aid use and hospitalization among Medicare beneficiaries with functional hearing loss a b .
| Any hospitalization c | Number of hospitalizations d | |||
|---|---|---|---|---|
| Variable | Odds ratio (95% CI) | Standard error | Incident rate ratio (95% CI) | Standard error |
| Self-reported functional hearing status | ||||
| Hearing aid user | Ref. | Ref. | ||
| A little trouble hearing, no hearing aid | 1.15 (0.92–1.44) | 0.13 | 1.10 (0.86–1.40) | 0.13 |
| A lot of trouble hearing, no hearing aid | 1.59 (1.18–2.16) * | 0.24 | 1.57 (0.97–2.53) | 0.38 |
| Age (y) | 1.02 (1.01–1.04) * | 0.01 | 1.02 (1.00–1.03) * | 0.01 |
| Female | 0.99 (0.79–1.22) | 0.11 | 0.89 (0.71–1.12) | 0.10 |
| Race/Ethnicity | ||||
| Non-Hispanic white | Ref. | Ref. | ||
| Non-Hispanic black | 1.14 (0.78–1.67) | 0.21 | 1.07 (0.75–1.52) | 0.19 |
| Hispanic | 1.07 (0.70–1.63) | 0.13 | 1.15 (0.70–1.87) | 0.28 |
| Other | 1.16 (0.69–1.97) | 0.31 | 0.94 (0.55–1.60) | 0.25 |
| Education | ||||
| Less than high school | Ref. | Ref. | ||
| High school graduate | 1.11 (0.85–1.44) | 0.14 | 1.09 (0.82–1.45) | 0.16 |
| Vocational, some college, associates | 1.07 (0.80–1.44) | 0.16 | 1.13 (0.88–1.46) | 0.14 |
| Bachelors or higher degree | 0.74 (0.51–1.06) | 0.14 | 0.84 (0.55–1.27) | 0.18 |
| Income to poverty ratio | ||||
| Less than 100% | Ref. | Ref. | ||
| 101–200% | 0.84 (0.61–1.15) | 0.13 | 0.79 (0.53–1.20) | 0.16 |
| Greater than 200% | 0.76 (0.56–1.04) | 0.12 | 0.63 (0.44–0.90) * | 0.11 |
| Chronic comorbidities count e | ||||
| 0 | Ref. | Ref. | ||
| 1–2 | 3.45 (1.87–6.37) ** | 1.06 | 4.30 (2.33–7.93) ** | 1.33 |
| 3–5 | 6.65 (3.57–12.38) ** | 2.08 | 8.74 (4.82–15.84) ** | 2.62 |
| ≥ 6 | 7.22 (3.73 - 14.00) ** | 2.41 | 9.22 (4.89–17.36) ** | 2.94 |
p < 0.05.
p < 0.001.
Data derived from the 2016 Medicare Current Beneficiary Survey Cost and Use File.
Survey weights applied according to Medicare Current Beneficiary Survey; percentages are based on weighted survey sample.
Logistic regression model.
Negative binominal regression model.
Chronic comorbidities include cancer, chronic obstructive pulmonary disease/asthma, chronic heart disease, serious mental illness, acute myocardial infarction, hypertension, diabetes, stroke, depression, arthritis, and dementia or Alzheimer's disease.
Discussion
In a study of 12,654 Medicare beneficiaries, we found that individuals who self-reported a lot of trouble hearing, even with a hearing aid, had higher odds of experiencing any hospitalization in the past year. Moreover, results suggest Medicare beneficiaries with functional hearing loss who report it interferes with communication with their health care provider have greater odds of any hospitalization and a higher rate of hospitalization. We also found patients who reported no hearing aid use and a lot of trouble hearing had increased odds of hospitalization compared with those who used hearing aids. These results corroborate previous evidence of hearing loss and hospitalization in a nationally representative sample of Medicare beneficiaries and build upon previous studies by analyzing the patient's perceived difficulty communicating with health care providers due to hearing status and association with hospitalization. Hearing loss may be a modifiable variable associated with hospitalization, such that hearing aid use may prevent hospitalizations.
Our study found a significant association between the odds of self-reporting a lot of trouble hearing and experiencing a hospitalization; however, we did not find a significant association between the degree self-reported functional hearing loss and rate of hospitalization in the past year. Notably, the CIs in the negative binomial regression model ( Table 2 ) did not cross 0 by a large margin. It is possible that a significant association would have been found with a larger sample size. Previous work in a large claims database of the general population of adults 50 years and older in the United States found an association between hearing loss identified by claims and rate of hospitalization. Specifically, this study reported incident rate ratios were 1.20, 1.30, and 1.47 over 2-, 5-, and 10-year time periods, respectively. 16 Similarly, Genther et al reported higher rate of hospitalization among those with hearing loss in the Health, Aging and Body Composition study. 21 In the current study, the functional nature of the self-report trouble hearing, with a hearing aid if appropriate, resulted in some adults with hearing loss (i.e., wear a hearing aid), but who consider themselves functionally without trouble, being allocated to the reference group which may overestimate the results. This may explain the large incident rate ratio (1.47) in a short time span (1 year) compared with other studies. In addition, hearing and odds of any hospitalization have previous been associated in studies using hearing measured per standard clinical protocols (i.e., pure tone audiometry). In one analysis of samples from the National Health and Nutrition Examination Survey (NHANES), authors found similar odds (OR: 1.35, 95% CI: 1.09–1.68) of any hospitalization to Medicare beneficiaries who report a lot of trouble in the current study. 15
In a second analysis, self-report communication difficulty with physicians due to hearing trouble was found to be associated with odds of hospitalizations. These findings align with previous findings in the 2013 MCBS data that individuals who reported trouble communicating with their doctor had 32% greater odds of 30-day hospital readmission. 22 Others have hypothesized that communication may be a key mediator in the association of hearing and health care outcomes. Adults with hearing loss are more likely to report poorer patient–provider communication. 23 It is plausible that poor communication due to hearing loss could impact treatment understanding and previous poor experiences (i.e., communication breakdowns) and deter adults with hearing loss from seeking preventative care in the future. A lack of preventative care could lead to an increase in hospitalizations.
In a third model, hearing aid use was found to be protective relative to not using a hearing aid among those with self-report trouble hearing. Specifically, participants who experienced a lot of trouble hearing and did not use hearing aids had greater odds of being hospitalized compared with those who are hearing aid users. A previous study using Medicare Expenditure Panel Survey data reported that hearing aid use reduced the probability of emergency department visits and hospitalizations, as well as the number of nights in the hospital. 18 The improved auditory signal from hearing aids may contribute to improved communication in health care settings. However, these results should be interpreted with extreme caution. Hearing aid use is associated with factors such as higher income and more education 24 that are protective of poor health care outcomes. It is difficult in secondary analyses to decouple these factors from the analysis. Prospective research in randomized trials is needed to confirm findings.
These findings have health, educational, clinical, and policy implications. As the body of research continues to grow, it is increasingly clear that adults with hearing loss are a vulnerable population who experience poorer health care utilization and incur increased expenditures relative to those without hearing loss. Health education programming should consider integrating this into courses to improve awareness in health care providers. 25 Clinically, sustainable programs assessing and addressing hearing loss to improve patient–provider communication should be pursued. Approaches that consider the environment, communication training for providers, and technology could mitigate the association between hearing and health care outcomes. 26 Such interventions should consider sustainability principles from implementation science to ensure their continued use and success. Moreover, the COVID-19 pandemic has offered insight into a health care system that increasingly relies on remote health care tele/video conferencing. Clinical models should consider optimizing these settings for adults with hearing loss as well. From a policy perspective, research assessing the impact of hearing care on health care utilization and the cost–benefit of such care should be closely monitored. Medicare could consider hearing aid coverage. Notably, should high-quality research prove hearing care improves overall health care utilization and spending, it may prove that the cost of hearing care is less than the overall savings in health care spending it could create. Scientifically rigorous methods, such as randomized control trials, are needed to assess this association. Audiologists are a prime group to champion these changes in health care education, clinical models to address hearing loss, and policy changes regarding hearing care under Medicare.
Limitations
This study used the Medicare Current Beneficiary Survey in which variables were self-reported. Self-reported measures of hearing may underestimate the prevalence of hearing loss. The misclassification of participants who have some degree of hearing loss into the no hearing loss category could impact the reported association and result in some bias. Moreover, the accuracy of proxy-reported measures of hearing and communication is unknown. Furthermore, the reported ownership of hearing aids does not mean the individuals are utilizing them on a daily basis or during health care interactions. Lastly, unmeasured residual confounding could impact the association. Further work is needed using clinical measure of hearing over a longitudinal time period to continue to define the association between hearing and hospitalization.
Conclusion
The present study found an association between self-report functional hearing loss and odds of any hospitalization in a nationally representative sample of Medicare beneficiaries. Importantly, self-reported difficulty communicating with health care providers due to hearing was associated with increased odds of any hospitalization and higher rate of hospitalization. Notably, nearly one in five Medicare beneficiaries with functional hearing loss said that it impacts on their communication with their health care provider. Findings suggest a lack of hearing aid use may increase rate of hospitalization. However, further work is needed. Given the high rates of hearing loss among older adults, a greater awareness of hearing loss among providers and policy workers and the way it may be negatively impact health utilization is needed.
Conflict of Interest None declared.
Equally contributed to this article.
References
- 1.Goman A M, Lin F R. Prevalence of hearing loss by severity in the United States. Am J Public Health. 2016;106(10):1820–1822. doi: 10.2105/AJPH.2016.303299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lin F R, Niparko J K, Ferrucci L. Hearing loss prevalence in the United States. Arch Intern Med. 2011;171(20):1851–1852. doi: 10.1001/archinternmed.2011.506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Goman A M, Reed N S, Lin F R. Addressing estimated hearing loss in adults in 2060. JAMA Otolaryngol Head Neck Surg. 2017;143(07):733–734. doi: 10.1001/jamaoto.2016.4642. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Health ABC Study Group . Lin F R, Yaffe K, Xia J. Hearing loss and cognitive decline in older adults. JAMA Intern Med. 2013;173(04):293–299. doi: 10.1001/jamainternmed.2013.1868. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Shukla A, Cudjoe T KM, Lin F R, Reed N S. Functional hearing loss and social engagement among Medicare beneficiaries. J Gerontol B Psychol Sci Soc Sci. 2021;76(01):195–200. doi: 10.1093/geronb/gbz094. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Dalton D S, Cruickshanks K J, Klein B E, Klein R, Wiley T L, Nondahl D M. The impact of hearing loss on quality of life in older adults. Gerontologist. 2003;43(05):661–668. doi: 10.1093/geront/43.5.661. [DOI] [PubMed] [Google Scholar]
- 7.Deal J A, Albert M S, Arnold M. A randomized feasibility pilot trial of hearing treatment for reducing cognitive decline: results from the Aging and Cognitive Health Evaluation in Elders Pilot Study. Alzheimers Dement (N Y) 2017;3(03):410–415. doi: 10.1016/j.trci.2017.06.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Institute of Medicine Committee on Quality of Health Care in America . Washington, DC: National Academies Press (US); 2001. Crossing the Quality Chasm: A New Health System for the 21st Century. [Google Scholar]
- 9.Shukla A, Nieman C L, Price C, Harper M, Lin F R, Reed N S. Impact of hearing loss on patient–provider communication among hospitalized patients: a systematic review. Am J Med Qual. 2019;34(03):284–292. doi: 10.1177/1062860618798926. [DOI] [PubMed] [Google Scholar]
- 10.Makoul G. Essential elements of communication in medical encounters: the Kalamazoo consensus statement. Acad Med. 2001;76(04):390–393. doi: 10.1097/00001888-200104000-00021. [DOI] [PubMed] [Google Scholar]
- 11.Zolnierek K B, Dimatteo M R. Physician communication and patient adherence to treatment: a meta-analysis. Med Care. 2009;47(08):826–834. doi: 10.1097/MLR.0b013e31819a5acc. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Clever S L, Jin L, Levinson W, Meltzer D O.Does doctor-patient communication affect patient satisfaction with hospital care? Results of an analysis with a novel instrumental variable Health Serv Res 200843(5, Pt 1):1505–1519. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Stein S M, Day M, Karia R, Hutzler L, Bosco J A., III Patients' perceptions of care are associated with quality of hospital care: a survey of 4605 hospitals. Am J Med Qual. 2015;30(04):382–388. doi: 10.1177/1062860614530773. [DOI] [PubMed] [Google Scholar]
- 14.Cudmore V, Henn P, O'Tuathaigh C MP, Smith S. Age-related hearing loss and communication breakdown in the clinical setting. JAMA Otolaryngol Head Neck Surg. 2017;143(10):1054–1055. doi: 10.1001/jamaoto.2017.1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Genther D J, Frick K D, Chen D, Betz J, Lin F R. Association of hearing loss with hospitalization and burden of disease in older adults. JAMA. 2013;309(22):2322–2324. doi: 10.1001/jama.2013.5912. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Reed N S, Altan A, Deal J A. Trends in health care costs and utilization associated with untreated hearing loss over 10 years. JAMA Otolaryngol Head Neck Surg. 2019;145(01):27–34. doi: 10.1001/jamaoto.2018.2875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Reed N S, Betz J F, Kucharska-Newton A M, Lin F R, Deal J A. Hearing loss and satisfaction with healthcare: an unexplored relationship. J Am Geriatr Soc. 2019;67(03):624–626. doi: 10.1111/jgs.15689. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Mahmoudi E, Zazove P, Meade M, McKee M M. Association between hearing aid use and health care use and cost among older adults with hearing loss. JAMA Otolaryngol Head Neck Surg. 2018;144(06):498–505. doi: 10.1001/jamaoto.2018.0273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Adler G S. A profile of the Medicare current beneficiary survey. Health Care Financ Rev. 1994;15(04):153–163. [PMC free article] [PubMed] [Google Scholar]
- 20.Aday L A, Andersen R. A framework for the study of access to medical care. Health Serv Res. 1974;9(03):208–220. [PMC free article] [PubMed] [Google Scholar]
- 21.Health, Aging and Body Composition Study . Genther D J, Betz J, Pratt S. Association between hearing impairment and risk of hospitalization in older adults. J Am Geriatr Soc. 2015;63(06):1146–1152. doi: 10.1111/jgs.13456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chang J E, Weinstein B, Chodosh J, Blustein J. Hospital readmission risk for patients with self-reported hearing loss and communication trouble. J Am Geriatr Soc. 2018;66(11):2227–2228. doi: 10.1111/jgs.15545. [DOI] [PubMed] [Google Scholar]
- 23.Mick P, Foley D M, Lin F R. Hearing loss is associated with poorer ratings of patient-physician communication and healthcare quality. J Am Geriatr Soc. 2014;62(11):2207–2209. doi: 10.1111/jgs.13113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nieman C L, Marrone N, Szanton S L, Thorpe R J, Jr, Lin F R. Racial/ethnic and socioeconomic disparities in hearing health care among older Americans. J Aging Health. 2016;28(01):68–94. doi: 10.1177/0898264315585505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Smith S, Manan N SIA, Toner S. Age-related hearing loss and provider-patient communication across primary and secondary care settings: a cross-sectional study. Age Ageing. 2020;49(05):873–877. doi: 10.1093/ageing/afaa041. [DOI] [PubMed] [Google Scholar]
- 26.Reed N S, Ferrante L E, Oh E S. Addressing hearing loss to improve communication during the COVID-19 pandemic. J Am Geriatr Soc. 2020;68(09):1924–1926. doi: 10.1111/jgs.16674. [DOI] [PMC free article] [PubMed] [Google Scholar]