Traumatic Brain Injury and Hearing Loss among older Medicare Beneficiaries

INTRODUCTION

Traumatic brain injury (TBI) is a common fall-related injury among older adults.¹ TBI results in significant disability among older adults.² Multiple studies have observed an association between TBI and hearing loss or auditory dysfunction, but few have examined older adults specifcally.^3,4

The association between TBI and increased risk for Alzheimer’s disease and related dementias (ADRD) is well documented, although mechanisms underlying this association are not well understood.⁵ One underexplored mechanism may be hearing loss, a modifiable risk factor for ADRD that is also commonly observed after TBI.³ Older individuals with hearing loss are 40% more likely to develop ADRD.⁶

Given the limited understanding of hearing loss following TBI among older adults, known associations with ADRD, and the value patients place on maintaining hearing function through the lifespan, the objective of the present study was to evaluate the incidence of new hearing loss diagnoses following TBI among older adults and secondarily to identify patient characteristics associated with new hearing loss diagnoses.

METHODS

We conducted a retrospective cohort study using Medicare administrative claims data obtained from the Center for Medicaid and Medicare Services Chronic Conditions Warehouse. Administrative claims from a 5% random sample for years 2010–2016 and a 20% random sample for years 2017–2018 were used for analysis.

Study Population

Community-dwelling Medicare beneficiaries aged 65 and older hospitalized with a TBI between 2010–2017 were eligible for the study. TBI was defined using the Centers for Disease Control and Prevention International Classification of Disease version 9 and version 10 surveillance codes on inpatient claims in the primary position and the date of TBI was assigned as the index date.¹

Hearing Loss

Hearing loss was defined based on prior work using ICD-9 and ICD-10 codes for problems with hearing, presbycusis, hearing loss, sensorineural hearing loss, and mixed conductive and sensorineural hearing loss (Supplemental Table S1).⁷ We excluded individuals with hearing loss diagnoses during the six months prior to TBI from analyses.

Statistical Analysis

We calculated follow-up from the date of TBI to the date of hearing loss diagnosis, death, or end of study period (i.e. 365 days), whichever came first. We divided the number of new hearing loss diagnoses (up to one diagnosis per beneficiary was counted) by the total follow-up time, and calculated 95% confidence intervals using standard methods for rates.

To identify factors associated with new hearing loss diagnosis following TBI, we used Cox proportional hazards models. We modeled time to new hearing loss diagnosis in the presence of death as a competing risk given significant censoring post-TBI using backwards selection.

This study was determined to be exempt from review by the Institutional Review Board at the University of Maryland School of Medicine.

RESULTS

Of 18,138 beneficiaries meeting study criteria, 2,030 (11.2%) were newly diagnosed with hearing loss (Table 1). Incidence of new hearing loss diagnosis over the year post-TBI was 14.1 (95% confidence interval (CI) 13.5, 14.8) per 100 person-years.

Table 1.

Characteristics of Participants with Traumatic Brain Injury (TBI) by Hearing Loss Diagnosis Status After TBI. Those w/hearing loss pre-TBI excluded, n=18,138

Mean (SD) or n (%)	No post-TBI hearing loss, n=16,108		Post-TBI hearing Loss, n=2,030		p-value
Age in years	81.1	(8.1)	82.3	(8.0)	<.001
Sex					.007
Female	9,210	(57.2)	1,097	(54.0)
Male	6,898	(42.8)	933	(46.0)
Race/Ethnicity					.002
White	14,182	(88.0)	1,834	(90.3)
Black	849	(5.3)	68	(3.4)
Asian	370	(2.3)	48	(2.4)
Hispanic	282	(1.8)	26	(1.3)
Other	425	(2.6)	54	(2.7)
Dually Eligible for Medicaid	4,317	(26.8)	390	(19.2)	<.001
OREC1 Age	14,342	(89.0)	1,847	(91.0)	.008
Injury-related Variables
AIS2 Head					<.001
Mild TBI (1–2)	2,405	(14.9)	269	(13.3)
Moderate TBI (3)	4,951	(30.7)	726	(35.8)
Severe TBI (4–5)	8,624	(53.5)	1,008	(49.7)
Length of Hospital Stay					.10
<2 days	2,291	(14.2)	254	(12.5)
≥2 or <4 days	5,747	(35.7)	709	(34.9)
≥4 or <6 days	3,583	(22.2)	479	(23.6)
≥6 days	4,487	(27.9)	588	(29.0)
Discharge to Home	8,120	(50.4)	944	(46.5)	<.001
Comorbidities
ADRD3	5,414	(33.6)	598	(29.5)	<.001
Anemia	11,604	(72.0)	1,506	(74.2)	.04
Asthma	2,552	(15.80)	366	(18.00)	.01
Cataract	12,545	(77.90)	1,701	(83.80)	<.001
Chronic Kidney Disease	6,847	(42.5)	864	(42.6)	.96
COPD4	5,783	(35.9)	708	(34.9)	.36
Depression	7,430	(46.1)	949	(46.8)	.60
Diabetes	7,291	(45.3)	923	(45.5)	.86
Glaucoma	4,567	(28.4)	673	(33.2)	<.001
Heart Failure	7,158	(44.4)	881	(43.4)	.37
Hyperlipidemia	13,743	(85.3)	1,801	(88.7)	<.001
Hyperplasia	3,979	(24.7)	590	(29.1)	<.001
Hypertension	14,643	(90.9)	1,846	(91.0)	.91
Hypothyroid	5,344	(33.20)	721	(35.50)	.04
Ischemic Heart Disease	10,818	(67.2)	1,408	(69.4)	.05
RA/OA5	11,208	(69.6)	1,482	(73.0)	.002

¹orginal reason for entitlement code (age vs. disability or end stage renal disease);

²abbreviated injury scale;

³Alzheimer’s disease and related dementias;

⁴chronic obstructive pulmonary disease;

⁵rheumatoid or osteoarthritis

Relative to men, women were less likely to receive a new diagnosis of hearing impairment post-TBI (hazard ratio (HR) 0.74; 95% CI 0.67, 0.81)(Figure and Supplementary Table S2). Several comorbidities (e.g. depression, hyperlipidemia) were associated with increased likelihood of new hearing loss diagnosis while ADRD was associated with decreased diagnosis of hearing loss (HR 0.74; 95% CI 0.67, 0.81). Relative to beneficiaries who are White, those who are Black were less likely to receive a diagnosis of hearing loss (HR 0.71; 95% CI 0.55, 0.91).

Figure 1.

Independent predictors of incident diagnosis of hearing loss following traumatic brain injury (TBI) with death as competing risk, n=18,138

Reference categories: Discharge home vs. other locations; moderate (abbreviated injury scale score (AIS) 3 and severe TBI (AIS 4–5) vs. mild TBI (AIS 1–2), female vs. male sex. ADRD: Alzheimer’s disease and related dementias

DISCUSSION

Among older Medicare beneficiaries hospitalized with TBI, incidence of new hearing loss diagnosis was 14.1/100 person-years. Prevalence of any hearing loss based on objective auditory assessments has been reported to range from 35%−73% among adults aged 65 and older, with 10%−36% experiencing moderate-severe hearing loss.⁸ These numbers suggest that the true number of older adults with new hearing loss post-TBI is likely higher than reported here.³

Individuals less likely to receive a hearing loss diagnosis following TBI were those who were younger, female, and Black. These findings are consistent with objective audiometric assessments conducted in the National Health and Aging Trends Study and the National Health and Nutrition Examination Survey.⁸

Diagnosis of ADRD was associated with decreased likelihood of a hearing loss diagnosis. This finding may represent a disparity in receipt of adequate care among older adults with ADRD⁹, particularly those who experience TBI. Importantly, hearing loss is increasingly recognized as a target for strategies to reduce risk for dementia and improve clinical outcome among those with ADRD.¹⁰ For older adults with TBI, screening for and treating hearing loss may represent an important avenue for ADRD risk reduction.

In this population of older adults hospitalized with TBI, incidence of new hearing loss diagnosis was 14%. Hearing loss post-TBI may represent a target for future research aimed at reducing risk for ADRD.

Supplementary Material

Supinfo

Supplementary Table S1. International Classification of Disease, Versions 9 (ICD-9) and 10 (ICD-10), Codes for Hearing Loss

Supplementary Table S2. Independent predictors of incident diagnosis of hearing loss following traumatic brain injury (TBI) with death as competing risk, n=18,138