The Effect of Near Universal Insurance Coverage on Use of Vision Care and Vision Among Adults with Diabetes: Evidence from Medicare Eligibility

Approximately 29 million individuals in the United States had diabetes in 2012, a figure which has grown over time. Diabetes is more prevalent among the elderly (26 percent in 2012) compared to other age groups (four percent and 16 percent of 20-44 year-olds and 45-64 year-olds, respectively, in 2012). Among other complications, diabetes is associated with various conditions that may adversely affect vision including diabetic retinopathy, cataracts, glaucoma, and age-related macular degeneration. Diabetic retinopathy, which affects approximately 30 percent of elderly adults with diabetes, is a leading cause of new cases of legal blindness.

While traditional Medicare does not cover preventive vision care for most beneficiaries, the program does provide coverage for those with diabetes. By lowering the out-of-pocket cost of vision care, Medicare coverage could increase use of vision care and improve outcomes among these individuals. Using 2000-2014 data from the National Health Interview Survey and a sample of 55-75 year-olds with self-reported diabetes, we examine whether there is a discontinuity in the likelihood of a recent eye care visit and self-reported visual acuity at age 65, which is the age at which most US adults become eligible for Medicare.

Our preliminary findings suggest that the likelihood of a recent eye care visit increases by about four percentage points (p<0.05) and the likelihood of needing but not purchasing eyeglasses or contacts due to cost decreases by about two percentage points (p<0.05) at age 65 among individuals with self-reported diabetes. These effects represent changes of about 7 percent and 19 percent, respectively, relative to the mean among 55-64 year-olds with self-reported diabetes. Further analyses will examine these and other outcomes by gender and race/ethnicity.

Since individuals with undiagnosed diabetes may be more likely to use medical care after gaining Medicare coverage, diagnosis of diabetes could increase discontinuously at age 65, which could confound our estimates. We do not find evidence of a discontinuous increase in self-reported diabetes at age 65 among all individuals aged 55-74. Nonetheless, we also limited our sample to those who reported that they were first diagnosed with diabetes prior to age 65 in some analyses. The estimated effects for this restricted sample were very similar to our main results.

Several studies have found a positive correlation but few have provided a causal link between vision insurance and use of vision care. The RAND health insurance experiment, a randomized study of cost sharing conducted more than 30 years ago, found evidence that reduced cost-sharing for vision services increased use of vision care and improved visual acuity outcomes. In fact, visual acuity was one of the few health outcomes that responded to more generous insurance coverage in this study. More recent work has examined the effects of Medicaid adult vision coverage on outcomes. However, this study is the first to our knowledge to examine individuals with diabetes, a group that is particularly vulnerable to vision problems.