Eye Care in the United States: Do We Deliver to High-Risk People Who Can Benefit Most From It?

Objective  To estimate the levels of self-reported access of eye care services in the nation.

Methods  We analyzed data from the 2002 National Health Interview Survey (30 920 adults aged ≥18 years). We estimated the number of US adults at high risk for serious vision loss and assessed factors associated with the use of eye care services.

Results  An estimated 61 million adults in the United States were at high risk for serious vision loss (they had diabetes, had vision or eye problems, or were aged ≥65 years); 42.0% of the 78 million adults who had dilated eye examinations in the past 12 months were among this group. Among the high-risk population, the probability of having a dilated eye examination increased with age, education, and income (P<.01). The probability of receiving an examination was higher for the insured, women, persons with diabetes, and those with vision or eye problems (P<.01). Approximately 5 million high-risk adults could not afford eyeglasses when needed; being female, having low income, not having insurance, and having vision or eye problems were each associated with such inability (P<.01).

Conclusions  There is substantial inequity in access to eye care in the United States. Better targeting of resources and efforts toward people at high risk may help reduce these disparities.

In 2000, approximately 3.3 million American adults aged 40 years or older were visually impaired (best-corrected visual acuity <20/40 in the better-seeing eye) mainly because of potentially preventable or treatable eye diseases such as age-related macular degeneration, cataract, diabetic retinopathy, or open-angle glaucoma.¹ Additionally, more than 11 million Americans aged 12 years and older needed refractive correction.² Visual impairment and eye diseases are associated with increased morbidity³ and mortality^4-13 as well as decreased quality of life.^14-17 These problems affect people's activities of daily living,^17-19 cause falls and injuries,^20-22 and lead to depression and social isolation.^23-26

By 2020, the number of people with visual impairment and eye diseases could increase by 50% or more.^1,27-30 From a patient's perspective, access to eye care is increasingly attached to the use of eye care services. Not surprisingly, use of—and thus “realized” or actual access to—eye care services is uneven in the United States, a special concern because of the growing demand for vision and eye care with a growing older population, the asymptomatic nature of many eye diseases in their early, treatable stages, and the prevalence of visual problems such as self-reported trouble or difficulty seeing. In this study, we estimated the levels of realized access to eye care at the national level and explored existing disparities and potential barriers to care. We paid special attention to the population at high risk for serious vision loss (persons with self-reported diabetes, with vision or eye problems, or aged ≥65 years), where preventive or regular eye care services are recommended.^31-33

We used vision data from the 2002 National Health Interview Survey (NHIS). Conducted annually by the National Center for Health Statistics, the NHIS uses a stratified, cross-sectional, multistage probability sample to make estimates for the civilian noninstitutionalized population. The methods of the NHIS have been described previously.^34,35 Briefly, the 2002 NHIS data were collected continuously through a personal household interview in the 50 states and the District of Columbia. Families in interviewed households provided basic health and demographic information for all members of the family. One adult and 1 child were randomly selected from each family for a more detailed health profile. In 2002, household interviews were completed for 93 386 persons living in 36 831 families who represented 36 161 households.³⁶ We restricted our analysis to the adult sample (n = 31 044). The conditional response rate of sample adults was 84.4%, and their final response rate was 74.4%.³⁷ In this study, we analyzed data from 30 920 adults aged 18 years and older using a well-established behavioral access model to guide the selection of variables.³⁸ Respondents who were blind or unable to see at all were not included in the analysis (n = 124).

Dependent variables in the study included respondent self-report of having visited an eye doctor in the past 12 months, having a dilated eye examination in the past 12 months, and not being able to afford eyeglasses when needed. These measures were chosen because visits to the eye doctor and receipt of a dilated eye examination capture differential elements of health care use while inability to afford eyeglasses also addresses the issue of financial barriers and the impact of reduced visual acuity due to uncorrected refractive error on vision-related quality of life.

Independent variables included age (ages 18-44, 45-64, and ≥65 years), sex, race or ethnicity (white [white non-Hispanic], African American [black non-Hispanic], Hispanic, and others [others non-Hispanic]), marital status, education (<high school, high school, and >high school), income (poverty income ratio <1, 1≤poverty income ratio<2, and poverty income ratio ≥2), health insurance, diabetes, vision or eye problems, and region (Northeast, Midwest, South, and West). Insurance groupings were collapsed from more detailed categorizations used in the NHIS. Respondents who had only public insurance (such as Medicare, Medicaid, Indian Health Service, military insurance, or other public or government insurance) were grouped into the public category, and the private group was for those with private coverage only. Respondents with both public and private coverage were categorized as both, and those who reported no coverage were categorized as none. Additionally, as the insurance status of persons aged 65 years and older may be unique because of Medicare, we further stratified this cohort (those aged ≥65 years) by 3 categories with a method previously used by the National Center for Health Statistics and assessed the influence of Medicare on eye care use.³⁹ These categories were Medicare only, Medicare plus other public coverage, and private (with or without Medicare). For those between the ages of 18 and 64 years, the 3 insurance categories used were public only, private (with or without public), and uninsured.

By high risk, we sought to identify the following: (1) those who are aged 65 years or older and thus have a much greater prior probability of having a significant chronic ocular condition based on the results of population-based surveys; (2) those who have a systemic condition (self-reported diabetes mellitus excluding borderline or gestational diabetes) associated with a significantly greater risk of blindness⁴⁰; and (3) those who already have an ocular disease or condition (age-related macular degeneration, cataract, diabetic retinopathy, glaucoma, or eye injury) that, if undetected or untreated, could result in permanent vision loss or those who report visual symptoms of “trouble seeing even with glasses or contact lenses” or activity limitations due to vision problems. We included this group of self-reported visual symptoms and activity limitations because other studies have shown that symptoms such as trouble seeing or blurred vision are independently associated with greater decrements in activities of daily living and instrumental activities of daily living as well as lower general health states as measured by the 36-item Short-Form Health Survey.^41,42 As such, this captures a larger group of persons who might well benefit from interventions to prevent further vision loss and activity limitations.

To make estimates representative of the civilian noninstitutionalized population, we used SAS statistical software version 9.1 (SAS Institute, Cary, NC) and SUDAAN statistical software version 9.0 (Research Triangle Institute, Research Triangle Park, NC) to adjust for the complex design of the sample, the problem of nonresponse, and the probability of selection. We used Wald χ² tests to explore the bivariate relationships. Multivariate logistic regressions were used to estimate the probability of visiting an eye doctor, having a dilated eye examination, and being unable to afford eyeglasses when needed after controlling for all other independent variables. Predictive margins and their standard errors from logistic regression models were estimated using Taylor linearization.⁴³

Of the respondents, 16.0% were aged 65 years and older, 6.5% had diabetes, and 19.5% had vision or eye problems (Table 1). Compared with the low-risk group, those at high risk tended to have less education and not as much income (as expressed by the poverty income ratio).

Of the estimated 61 million adults in the United States classified as being at high risk for serious vision loss, we estimated that only half visited an eye doctor in the past 12 months and half had a dilated eye examination (Figure; we assume that these groups overlapped very much). By comparison, of the estimated 144 million people classified as not being at high risk, about a third visited an eye doctor and a third had a dilated eye examination (again, we assume great overlap). We found that of an estimated 78 million people with dilated eye examinations, 42.0% were at high risk and 58.0% were not at high risk. We also found that 1 in 12 persons at high risk and 1 in 25 not at high risk could not afford eyeglasses when needed.

Results from multivariate regression among the general population (Table 2) showed significant relationships between the use of eye care and various individual and systematic descriptors. Controlling for all other variables in the model, the probability of visiting an eye doctor and the probability of having a dilated eye examination increased with age and education (P<.01). In addition, having high income, having health insurance, being female, having diabetes, or having vision or eye problems increased the probability of visiting an eye doctor and having the eye examination (P<.01). Inability to afford eyeglasses was associated with being unmarried, being female, having low income, being without insurance, having diabetes, or having vision or eye problems (P<.01). Although people at high risk for serious vision loss were more likely than the general population to receive eye care services (visits, examinations), health disparities among the high-risk population remained clear and crucial (Table 3). In this population (unlike the population generally), those who were married were more likely to visit an eye doctor (P<.05) or have a dilated eye examination (P<.01) than those who were not married, again when controlling for all other variables in the model.

Among adults younger than 65 years, those with private or public insurance had a higher probability of visiting an eye doctor, having a dilated eye examination, and being able to afford eyeglasses than those without insurance (Table 4). In contrast, we did not find significant differences for visits and examinations between persons with public and private coverage, but there was a significant difference between these groups in being able to afford eyeglasses. Among persons aged 65 years and older, those with private coverage (with or without Medicare) were more likely to have had visits and examinations and were better able to afford eyeglasses than were those with Medicare only.

As the baby boomers age, ensuring access to chronic care and preventive services such as vision and eye care has become a major public health concern.^44-46 Appropriately, Healthy People 2010⁴⁷ includes 10 objectives for vision that can be used to monitor the burden of visual impairment and levels of eye care in the nation. Our study systematically explored access to eye care at the national level. We found that only 54.8% of adults with vision or eye problems and 62.9% of adults with diabetes had a dilated eye examination in the past 12 months. The rates of eye doctor visits and dilated eye examinations were higher among our high-risk group than in those who did not have those characteristics. Greater access to eye care services was also associated with age, sex, race or ethnicity, education, income, health insurance, and region. These results are supported by the findings in population-based studies such as the Beaver Dam Eye Study⁴⁸ or the Salisbury Eye Evaluation Project,⁴⁹ which also found greater rates of use among those with ocular conditions or vision problems, diabetes, and additional health insurance.

One of the 2 major goals set by Healthy People 2010⁴⁷ is to eliminate health disparities between the sexes and by race or ethnicity and socioeconomic level. In this study, we found considerable differences by sex in realized access to eye care services, with women faring better than men in visiting eye doctors and having dilated eye examinations. We also found that African Americans and Hispanic persons were significantly less likely to visit an eye doctor than were white persons, which calls for continued efforts to improve the access to eye care services among minority populations as found in other studies.^49-52 These findings from this national sample are consistent with those in an article⁵³ on the Medicare population with diabetes that found that men and African Americans used eye care services less.

Not surprisingly, we found that adults with higher education and those with greater incomes were relatively more likely to use eye care services. In addition, our findings suggest that the uninsured were much less likely to receive preventive eye care in the form of eye doctor visits or dilated eye examinations than were the insured, which is consistent with a recent article⁵⁴ on the use of other recommended preventive services. Using data from the 2000 Behavioral Risk Factor Surveillance System, which also relies on self-reports, Nelson et al⁵¹ obtained findings consistent with ours, as they found that uninsured persons with diabetes were less likely to have an annual dilated eye examination and to undergo other tests related to diabetes management. Our findings are also consistent with data from 5 states in the optional vision module offered by the Behavioral Risk Factor Surveillance System, which indicated that “no reason to go” and “cost or insurance” were the top 2 reasons for people not to see an eye care professional.⁵⁵ Therefore, in addition to the widely recognized issue of awareness regarding access to eye care,⁵⁶ we see that health insurance remains a critical factor.

Many conditions causing visual impairment and blindness are often asymptomatic in their early, treatable stages. Timely examination of vision is recommended for people with vision or eye problems, and a yearly examination is recommended for those with diabetes or at older ages (ages ≥65 years).^31-33 Screening the vision of adults aged 65 years and older is listed among the top 10 priorities for clinical preventive services (as determined by the National Commission on Prevention Priorities).⁵⁷ However, we found that not much more than half of those at high risk for serious vision loss had visited an eye doctor in the past 12 months; the estimate for a dilated eye examination in that period was similar. Of particular concern is that even though diabetes is the leading cause of legal blindness in the United States,³³ 1 in 3 adults with diabetes did not have a dilated eye examination in the preceding year, a finding that has remained consistent across numerous studies for many years.^51,58

Adults with vision or eye problems or with diabetes used significantly more eye care services than their counterparts without these problems. The age-adjusted prevalence of visual impairment due to refractive error was found to be higher among persons with diabetes than persons without diabetes (unpublished data, Centers for Disease Control and Prevention, 1999-2004). Even so, we found that 10.5% of persons with vision or eye problems and 8.8% of those with diabetes could not afford eyeglasses when needed as compared with 4.2% of persons without vision or eye problems and 5.2% of persons without diabetes. Thus, the people who needed the visual aid more were the ones less likely to be able to afford these aids. A recent article with data on the general population from the National Eye Institute² and newly released data on people with diabetes from the Centers for Disease Control and Prevention⁵⁹ both call for increased public health interventions in reducing the burden of correctable visual impairment due to refractive error among the general public and especially persons with diabetes.

There were several limitations to our study. First, our definition of high-risk population does not account for factors such as family history and genetic features; relying on this definition may have caused us to underestimate the number of high-risk adults. Alternative definitions of high risk would include characterization along racial or ethnic lines or by socioeconomic status. In our multivariate analyses, we included these as independent factors and found, as expected, that those known from prior studies to be at higher risk indeed had lower rates of visits to the eye doctor and dilated eye examinations. We included these in a multivariate analysis as opposed to subgroup analyses because these are sociodemographic characteristics already known to be associated with reduced eye care use. Our current study instead evaluates patients on the basis of known medical and ocular conditions as well as demographic characteristics (age) and patient-centered symptoms known to have higher levels of use and found not only that was there a substantial gap in care received but that the impact of known characteristics persisted in this select group who would most benefit from regular care.

Second, our study is based on self-reported data. Inevitably, there is some degree of difference between self-reported outcomes and clinical or measurable outcomes, but the survey results from the NHIS are generally considered valid and reliable.^34,35

Third, the institutionalized population (eg, residents of nursing homes) and military groups were not included in the survey. Also, there was some nonresponse, especially for income information. Still, population-based surveys are an effective way to systematically monitor and evaluate access to eye care, and the NHIS is considered one of the best available national samples and has a relatively high response rate, which increases its generalizability.

A fourth concern is our somewhat puzzling result that some estimates were higher for a dilated eye examination than for a visit to an eye doctor. Although the reasons behind this unexpected outcome are unknown and potentially complicated (eg, eye care services provided by a primary care physician, recall bias in examinations, inability to understand the survey questions, vision screening by people who would not be identified as eye doctors), we can speculate that awareness of what was going on during the visit(s) of interest may have been at the core of the problem. Among persons who had not finished high school, we found that the percentage for visiting an eye doctor was well below the percentage for receiving a dilated eye examination, but among those with more than a high school education, we found a marginal difference in the other direction. The definition of dilation in the wording of the question (“ . . . dilated? This would have made you temporarily sensitive to bright light”) likely contributed to this finding as well. Finally, we are concerned that although access to medical care is a multidimensional concept,^38,60-64 we could not explore information on many of the factors that might contribute to whether US adults obtain reasonable access to eye care. Owing to the limitations of the data set, we could not explore information on, inter alia, urban or rural residence, the characteristics of providers, and patient satisfaction. Future research might benefit by incorporating these measures.

To summarize, despite the recent emphasis on vision in Healthy People 2010,⁴⁷ there is still substantial inequity in realized access (ie, actual use) to eye care in the United States. Of the estimated 61 million adults in the United States classified as being at high risk for serious vision loss, we found that just slightly more than half had received a dilated eye examination in the past 12 months. We also found that 1 in 12 high-risk persons could not afford eyeglasses when needed. Better targeting of resources and efforts toward people at high risk for whom cost-effective interventions exist may help reduce these disparities.

Correspondence: Xinzhi Zhang, MD, PhD, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, 4770 Buford Hwy NE (K-10), Atlanta, GA 30341-3727 (xzhang4@cdc.gov).

Submitted for Publication: November 4, 2006; accepted December 6, 2006.

Author Contributions: Dr Zhang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Financial Disclosure: None reported.

Disclaimer: The findings and conclusions in this article are those of the authors and do not necessarily represent the views of the Centers for Disease Control and Prevention.