This cross-sectional population-based survey of Medicare beneficiaries 65 years and older examines if sociodemographic disparities are associated with low-vision devices and low-vision rehabilitation.
Key Points
Question
Are there sociodemographic disparities in the use of low-vision services among Medicare beneficiaries 65 years and older?
Findings
In this cross-sectional survey study of a nationally representative sample, older adults in the United States from minority racial/ethnic groups were less likely to report using low-vision devices but not vision rehabilitation compared with white individuals.
Meaning
If the associations from this study are confirmed, coverage of low-vision devices by Medicare may help to address a significant health care disparity in the use of this evidence-based intervention.
Abstract
Importance
Low-vision assistive devices are not covered by Medicare and many private insurers, although there is evidence that they can improve functioning and quality of life. Little is known about whether sociodemographic disparities exist in the use of low-vision services by Medicare beneficiaries.
Objective
To determine if sociodemographic disparities exist in the use of low-vision services by Medicare beneficiaries.
Design, Setting, and Participants
Cross-sectional population-based survey. The National Health Interview Survey is an annually administered nationally representative US survey. Data used in this study were from the 2002, 2008, and 2016 National Health Interview Survey vision supplement. Participants who were Medicare beneficiaries 65 years and older with self-reported vision impairment were included.
Main Outcomes and Measures
Multivariable logistic regression was performed to evaluate if sociodemographic or economic factors were associated with self-reported use of low-vision devices or low-vision rehabilitation among Medicare beneficiaries 65 years and older who self-reported vision impairment.
Results
There were 3058 participants included in the study. The survey weighted proportion of participants who were men was 37.9% (95% CI, 35.8%-40.0%), while 79.1% (95% CI, 77.2%-80.9%) were non-Hispanic white, 10.2% (95% CI, 9.0%-11.5%) were non-Hispanic black, 6.7% (95% CI, 5.6%-8.1%) were Hispanic, and 4.0% (95% CI, 3.2%-5.0%) identified with another race/ethnicity. The weighted proportion who reported using low-vison devices and low-vision rehabilitation were 26.1% (95% CI, 24.2%-28.1%) and 3.5% (95% CI, 2.8%-4.3%), respectively. In a model adjusted for ocular diagnoses, Hispanic individuals (odds ratio, 0.61; 95% CI, 0.39-0.97) and individuals from other races/ethnicities (odds ratio, 0.39; 95% CI, 0.19-0.80), but not black individuals, were significantly less likely to report using low-vision devices than white individuals. In the model that was not adjusted for ocular diagnoses, black individuals (odds ratio, 0.73; 95% CI, 0.54-0.99) were also significantly less likely to report using low-vision devices. There were no significant racial/ethnic disparities for reported use of low-vision rehabilitation.
Conclusions and Relevance
Additional research is needed to clarify the association between sociodemographics and use of low-vision services in the Medicare population. However, policy makers could consider expanding Medicare coverage to include low-vision devices in an effort to address significant disparities in the use of this evidence-based intervention.
Introduction
Low vision refers to uncorrectable vision impairment (VI). In 2017, more than 3.8 million adults older than 45 years in the United States were estimated to have low vision, and this number is expected to double by 2050.1 Impaired vision is associated with decreased quality of life and increased depression, falls, and mortality.2
Low-vision rehabilitation and assistive devices can improve functioning and quality of life.2,3 Vision rehabilitation includes evaluation of activities of daily living and implementation of a therapeutic plan to improve functioning. Components of vision rehabilitation may include mobility training, training in adaptive strategies, and environmental modification, among other interventions. Under current Medicare policy, vision rehabilitation is reimbursable if prescribed by a physician or optometrist and implemented by Medicare providers (eg, physical and occupational therapists).
In contrast, Medicare and many private insurers do not provide coverage for low-vision devices. Current Medicare guidelines are interpreted such that low-vision devices are classified under the same statute as eyeglasses, which are not routinely covered. However, the purpose of low-vision devices, to improve independence in activities of daily living, is identical to that of other assistive equipment that is covered under Medicare for individuals with other disabilities.4
Prior studies have demonstrated socioeconomic disparities in the use of low-vision devices,5,6 although neither study was representative of the US population. In our study, using nationally representative data, we sought to determine whether there are disparities in low-vision device use among older adults with Medicare in the United States, information that may be important in advocating for Medicare coverage of low-vision devices.
Methods
Data Source
The National Health Interview Survey (NHIS) is a cross-sectional study used to monitor the health of the US population and has been administered annually through a computer-assisted in-person interview since 1997. A vision supplement with questions related to ocular health was included in 2002, 2008, and 2016. Our study sample consisted of adults 65 years or older with Medicare insurance who responded to the vision supplement and answered yes to the question, “Do you have any trouble seeing, even when wearing glasses or contact lenses?” The University of Michigan institutional review board deemed this study exempt because it was a secondary data analysis involving publicly available data.
Variables
Respondents were classified as low-vision device users if they answered yes to the question, “Do you use any adaptive devices such as telescopic or other prescriptive lenses, magnifiers, large print or talking materials, CCTV, white cane, or guide dog?” Respondents were classified as low-vision rehabilitation users if they answered yes to the question, “Do you use any vision rehabilitation services, such as job training, counseling, or training in daily living skills and mobility?”
Statistical Analyses
The 2016 NHIS vision supplement was used to generate nationally representative estimates of low-vision device and rehabilitation use. For all other analyses, household, family, and sample adult files were merged within each year and then combined across the 3 years of the vision supplement (2002, 2008, and 2016). Since each survey year fell within a different sample design period, we treated each year as statistically independent according to NHIS guidelines. We used the Taylor series linearization method to estimate variance. Based on the complex survey design, we calculated the weighted proportions of participants for each covariate stratified by low-vision device and rehabilitation use. We report unadjusted P values from Pearson χ2 tests. We performed multivariable logistic regression to examine associations with 2 outcomes: low-vision device and rehabilitation use. We generated 2 sets of models; the first was adjusted for age, sex, race/ethnicity, income relative to the federal poverty level, self-reported ocular comorbidities (cataract, glaucoma, macular degeneration, diabetic retinopathy), and survey year; the second was adjusted for each of the aforementioned covariates except ocular comorbidities. All analyses were conducted using STATA/MP version 15 (StataCorp LP) and accounted for the complex design of NHIS.
Results
This study included 3058 participants. In terms of sociodemographic differences, low-vision device (Table 1) users were significantly older and more likely to be white than nonusers, while vision rehabilitation (eTable 1 in the Supplement) users were older than nonusers. Based on the weighted survey data, we estimated that overall in 2016 there were 1 556 822 (95% CI, 1 317 667-1 795 977) low-vision device users and 239 438 (95% CI, 153 053-325 824) vision rehabilitation users among Medicare beneficiaries 65 years or older with self-reported VI.
Table 1. Summary Statistics of Participant Characteristics by Low-Vision Device Use.
| Characteristic | Weighted Proportion (95% CI) | P Valuea | |
|---|---|---|---|
| Low-Vision Device (n = 812) | No Low-Vision Device (n = 2242) | ||
| Total | 26.1 (24.2-28.1) | 74.0 (72.0-76.0) | NA |
| Men | 36.2 (32.1-40.5) | 38.5 (36.1-40.9) | .35 |
| Age, y | |||
| 65-69 | 16.8 (13.9-20.1) | 26.6 (24.4-28.8) | <.001 |
| 70-74 | 17.7 (14.8-21.0) | 24.8 (22.6-27.2) | |
| 75-79 | 20.3 (17.2-23.9) | 17.9 (16.1-19.9) | |
| 80-84 | 19.8 (16.7-23.4) | 14.6 (13.0-16.4) | |
| ≥85 | 27.2 (21.9-29.1) | 16.1 (14.3-18.1) | |
| Race/ethnicity | |||
| White | 86.1 (83.1-88.6) | 76.6 (74.3-78.8) | <.001 |
| Black | 8.1 (6.2-10.6) | 10.9 (9.6-12.5) | |
| Hispanic | 3.9 (2.7-5.6) | 7.7 (6.3-9.5) | |
| Other | 1.9 (0.9-3.6) | 4.7 (3.8-5.9) | |
| Education | |||
| <High school | 28.0 (24.5-31.9) | 30.2 (27.8-32.6) | .34 |
| High school graduate | 32.7 (28.7-36.9) | 28.8 (26.5-31.2) | |
| Some college | 22.8 (19.5-26.5) | 22.7 (20.6-24.9) | |
| ≥Bachelor’s degree | 16.5 (13.5-20.0) | 18.4 (16.5-20.4) | |
| Federal poverty level | |||
| ≤200 | 39.7 (35.8-43.7) | 42.1 (39.5-44.7) | .23 |
| 200-399 | 39.1 (35.2-43.1) | 34.9 (32.4-37.5) | |
| ≥400 | 21.3 (17.9-25.1) | 23.0 (21.0-25.2) | |
| Eye conditions | |||
| Diabetic retinopathy | 8.5 (6.4-11.2) | 4.0 (3.2-5.1) | <.001 |
| Cataract | 73.9 (70.0-77.4) | 66.8 (64.5-69.1) | .002 |
| Glaucoma | 17.7 (14.8-20.9) | 16.0 (14.2-17.9) | .34 |
| Macular degeneration | 30.7 (27.1-34.6) | 13.5 (11.8-15.3) | <.001 |
| Low-vision rehabilitation | 7.1 (5.3-9.4) | 2.2 (1.6-3.0) | <.001 |
Abbreviation: NA, not applicable.
P values are unadjusted and calculated using design-adjusted Pearson χ2 test.
In the model adjusted for all covariates including ocular diagnoses, the odds of using a low-vision device were significantly lower for Hispanic individuals (adjusted odds ratio [AOR], 0.61; 95% CI, 0.41-0.91) and individuals of other races/ethnicities (AOR, 0.39; 95% CI, 0.19-0.78) (Table 2). There was no significant difference in the likelihood of reporting vision rehabilitation use between racial/ethnic groups. In the model adjusted for all covariates except ocular diagnoses, there were also no racial/ethnic disparities in rehabilitation use, but black individuals (AOR, 0.73; 95% CI, 0.54-0.99), Hispanic individuals (AOR, 0.53; 95% CI, 0.36-0.78), and individuals of other races/ethnicities (AOR, 0.42; 95% CI, 0.22-0.80) were all less likely to report using low-vision devices (eTable 2 in the Supplement). Reporting use of a low-vision device was not significantly associated with sex, educational attainment, or income as a percent of the federal poverty level in any of our models.
Table 2. Logistic Regression Models for Low-Vision Device and Rehabilitation Use Among Adults With Vision Impairment 65 Years and Older With Medicare.
| Characteristic | Low-Vision Device Use | Low-Vision Rehabilitation Use | ||
|---|---|---|---|---|
| AOR (95% CI) | P Value | AOR (95% CI) | P Value | |
| Sex | ||||
| Male | 0.89 (0.71-1.11) | .31 | 0.83 (0.50-1.41) | .50 |
| Female | 1 [Reference] | NA | 1 [Reference] | NA |
| Race/ethnicity | ||||
| White | 1 [Reference] | NA | 1 [Reference] | NA |
| Black | 0.84 (0.60-1.19) | .34 | 1.15 (0.56-2.33) | .71 |
| Hispanic | 0.61 (0.39-0.97) | .04 | 1.29 (0.56-2.96) | .55 |
| Other | 0.39 (0.19-0.80) | .01 | 0.38 (0.09-1.61) | .19 |
| Education | ||||
| <High school | 0.90 (0.65-1.27) | .56 | 0.37 (0.16-0.85) | .02 |
| High school graduate | 1.09 (0.78-1.54) | .61 | 0.78 (0.38-1.61) | .50 |
| Some college | 1.05 (0.74-1.48) | .78 | 0.80 (0.39-1.64) | .54 |
| ≥Bachelor’s degree | 1 [Reference] | NA | 1 [Reference] | NA |
| Federal poverty level, % | ||||
| ≤200 | 1.01 (0.75-1.37) | .93 | 1.60 (0.83-3.12) | .16 |
| 200-399 | 1.15 (0.87-1.52) | .32 | 1.07 (0.54-2.12) | .85 |
| ≥400 | 1 [Reference] | NA | 1 [Reference] | NA |
| Eye conditions | ||||
| Diabetic retinopathy | 2.65 (1.72-4.09) | <.001 | 1.70 (0.67-4.34) | .27 |
| Cataract | 1.13 (0.90-1.42) | .30 | 0.78 (0.47-1.29) | .33 |
| Glaucoma | 1.04 (0.78-1.37) | .80 | 0.97 (0.55-1.70) | .91 |
| Macular degeneration | 2.27 (1.77-2.90) | <.001 | 1.76 (1.07-2.91) | .03 |
| Year | ||||
| 2016 | 0.67 (0.52-0.86) | .002 | 1.16 (0.66-2.07) | .60 |
| 2008 | 0.67 (0.51-0.89) | .004 | 1.25 (0.70-2.23) | .45 |
| 2002 | 1 [Reference] | NA | 1 [Reference] | NA |
Abbreviations: AOR, adjusted odds ratio; NA, not applicable.
Discussion
We found significant racial/ethnic disparities in the use of low-vision devices but not vision rehabilitation among older adults with Medicare. Since 2002, Medicare has provided coverage of low-vision rehabilitation services but continues not to cover low-vision devices. This is despite evidence from recent clinical trials demonstrating the efficacy of low-vision devices alone and together with vision rehabilitation for improving vision-dependent functioning.3
We presented models both adjusted and unadjusted for self-reported ocular diagnoses. We believe that each of these conveys an important and distinct message. The model adjusted for ocular diagnoses aims to nullify the effect of different eye conditions on patterns of access to low-vision services. Specifically, conditions that tend to cause central vision loss (eg, macular degeneration) are often more amenable to improvement with low-vision devices than those that cause other impairments like peripheral vision loss (eg, glaucoma).2,7 However, glaucoma is a highly prevalent eye disease that is more common in black and Hispanic individuals.8,9 The model that is unadjusted for ocular diagnoses may unmask meaningful disparities in low-vision treatment options available to minority populations.
Respondents who self-reported VI in NHIS likely represent a heterogeneous group with varying levels of VI. However, we do not know whether certain groups who were not candidates for low-vision devices or rehabilitation were more likely to self-report VI. If this was the case, it could have biased our results, with the overreporting group(s) appearing less likely to receive low-vision services. The limited evidence available from prior studies shows that black individuals in the Salisbury Eye Evaluation Study were more likely to self-report good vision but have poor visual acuity.10
Limitations
There were several limitations to our study. Our findings may be limited by recall and social desirability biases, the small number of participants reporting vision rehabilitation, and that we were not able to determine actual need for low-vision devices. While this study supports an association between race/ethnicity and use of low-vision devices, we cannot determine causation from these data.
Conclusions
In other areas of medicine, expansion of insurance coverage has decreased racial disparities.11 Additional research is needed to clarify the relationship between sociodemographics and use of low-vision services in the Medicare population. However, if the findings from this study are confirmed, it may suggest that policy makers could consider coverage of low-vision devices under Medicare in an effort to address disparities and expand access to this evidence-based intervention.
eTable 1. Summary Statistics of Participant Characteristics by Low Vision Rehabilitation Use
eTable 2. Logistic Regression Models for Low Vision Device and Rehabilitation Use Among Visually Impaired Adults age 65 and Older with Medicare (not adjusted for ocular comorbidities)
References
- 1.Chan T, Friedman DS, Bradley C, Massof R. Estimates of incidence and prevalence of visual impairment, low vision, and blindness in the United States. JAMA Ophthalmol. 2018;136(1):12-19. doi: 10.1001/jamaophthalmol.2017.4655 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Binns AM, Bunce C, Dickinson C, et al. How effective is low vision service provision? a systematic review. Surv Ophthalmol. 2012;57(1):34-65. doi: 10.1016/j.survophthal.2011.06.006 [DOI] [PubMed] [Google Scholar]
- 3.Stelmack JA, Tang XC, Wei Y, et al. ; LOVIT II Study Group . Outcomes of the Veterans Affairs Low Vision Intervention Trial II (LOVIT II): a randomized clinical trial. JAMA Ophthalmol. 2016;135(2):96. doi: 10.1001/jamaophthalmol.2016.4742 [DOI] [PubMed] [Google Scholar]
- 4.Morse AR, Massof RW, Cole RG, et al. Medicare coverage for vision assistive equipment. Arch Ophthalmol. 2010;128(10):1350-1357. doi: 10.1001/archophthalmol.2010.228 [DOI] [PubMed] [Google Scholar]
- 5.Tomita MR, Mann WC, Fraas LF, Burns LL. Racial differences of frail elders in assistive technology. Assist Technol. 1997;9(2):140-151. doi: 10.1080/10400435.1997.10132305 [DOI] [PubMed] [Google Scholar]
- 6.Kaye HS, Yeager P, Reed M. Disparities in usage of assistive technology among people with disabilities. Assist Technol. 2008;20(4):194-203. doi: 10.1080/10400435.2008.10131946 [DOI] [PubMed] [Google Scholar]
- 7.Patodia Y, Golesic E, Mao A, Hutnik CM. Clinical effectiveness of currently available low-vision devices in glaucoma patients with moderate-to-severe vision loss. Clin Ophthalmol. 2017;11:683-687. doi: 10.2147/OPTH.S128425 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Varma R, Ying-Lai M, Francis BA, et al. ; Los Angeles Latino Eye Study Group . Prevalence of open-angle glaucoma and ocular hypertension in Latinos: the Los Angeles Latino Eye Study. Ophthalmology. 2004;111(8):1439-1448. doi: 10.1016/j.ophtha.2004.01.025 [DOI] [PubMed] [Google Scholar]
- 9.Friedman DS, Jampel HD, Muñoz B, West SK. The prevalence of open-angle glaucoma among blacks and whites 73 years and older: the Salisbury Eye Evaluation Glaucoma Study. Arch Ophthalmol. 2006;124(11):1625-1630. doi: 10.1001/archopht.124.11.1625 [DOI] [PubMed] [Google Scholar]
- 10.El-Gasim M, Munoz B, West SK, Scott AW. Discrepancies in the concordance of self-reported vision status and visual acuity in the Salisbury Eye Evaluation Study. Ophthalmology. 2012;119(1):106-111. doi: 10.1016/j.ophtha.2011.07.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Hayes SL, Riley P, Radley DC, McCarthy D. Reducing racial and ethnic disparities in access to care: has the Affordable Care Act made a difference? Issue Brief (Commonw Fund). 2017;2017:1-14. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Summary Statistics of Participant Characteristics by Low Vision Rehabilitation Use
eTable 2. Logistic Regression Models for Low Vision Device and Rehabilitation Use Among Visually Impaired Adults age 65 and Older with Medicare (not adjusted for ocular comorbidities)
