Implications of Neighborhood- and Patient-Level Factors for Eye Care

In the study by Ong and colleagues titled “The relationship between neighborhood-level social determinants of health (SDOH) and presenting characteristics for rhegmatogenous retinal detachments (RRDs),¹ the researchers highlight the association of the neighborhood and built environment on eye health. This important work, that has been previously understudied, found that neighborhood-level factors, including lower per capita income, a higher percentage of workers who drove to work, and worse area deprivation, were associated with worse severity of presenting RRD. Additionally, the authors reported patient-level factors associated with worse severity of presenting RRD that included having no insurance, having public insurance, and identifying as non-Hispanic Black.¹

RRD is an eye condition that requires a timely surgical intervention by a retina surgeon for optimal outcomes. Accessing such a specialist represents a complex journey through the medical system for all patients. However, additional barriers may be experienced by individuals with social risk factors, either at the patient-level or at the neighborhood-level. For example, patients with RRD may first present to the Emergency Department (ED) for care. In one study, the majority (76.6%) of people who presented to the ED with RRD were not repaired the same day; however, White patients had a three times higher odds of receiving same-day repair compared with patients who identify as other races (Odds Ratio: 3.85; 95% Confidence Interval: 1.75–8.47; p=0.001).² Many EDs do not have retina surgeons on staff or on call, and so many people then need to be referred to a retina surgeon after their visit to the ED. However, it has been reported that a quarter of patients presenting to the ED with an ophthalmic condition are loss to follow-up, with a higher rate of loss-to-follow-up among patients that have Medicaid or are uninsured compared with those with commercial insurance.³ Similarly, if a patient first sees an optometrist or comprehensive ophthalmologist for their vision concerns and they have an RRD, they also need to be referred to a retina surgeon for repair. Both the referral to the retina surgeon and the wait for available operating room time are additional necessary steps in this complex care process where delays can also occur. Ong and colleagues’ study outlines how both individual risk factors and neighborhood-level risk factors may ultimately influence the presentation of RRD. Delays between these different steps in the care process are likely influenced by each person’s social risk factors, ultimately leading people with worse social risk factors to present to the retina surgeon with worse vision.

A complex interplay exists between individuals and broader social determinants of health – such as the neighborhood and built environment in which they live.⁴ This interplay is ripe for scientific discovery in ophthalmology. The individual factors that influence a person’s ability to cope when neighborhood-level resources are not available creates insights into disparities in health outcomes. Interactions occur when a risk factor, such as neighborhood transportation access, and an outcome such as severity of RRD presentation, differs across a third variable, such as race.⁵ For example, Ong and colleagues found that an increase in the neighborhood-level percentage of workers who drove to work was associated with an increased odds of presenting with RRDs with worse visual acuity.¹ At first, this finding seems counter-intuitive. Shouldn’t a neighborhood where more people have cars make it easier for people to attend healthcare visits? Perhaps in neighborhoods where a high percentage of people drive to work are neighborhoods that are farther from public transit. People are dependent on personal vehicles for transportation. In that scenario, a higher income person who lives in this neighborhood owns a car or pays for a rideshare service. But a low income person in this neighborhood cannot afford a car or rideshare and, without good public transit, that lower income person has more difficulty attending a visit with an eye doctor. We posit that there is an interaction between the neighborhood and the person that affects the outcome. An interaction analysis within the statistical model – to see if the association of the neighborhood-level variable of the percent of workers who drive to work varies by the individual’s income, may affect the odds of worse outcomes from RRD.

There are new insights to be gained by assessing both patient-level and neighborhood-level interactions separately and together via interaction affects. We will see more complex insights into eye and vision health disparities, which can then inform more nuanced policies to address these problems. In another study from the same tertiary referral center, Tang and colleagues found a interaction between increased neighborhood deprivation and race/ethnicity that was associated with the probability of lapses in diabetic retinopathy care.⁶ White patients who lived in a neighborhood with higher deprivation were more likely to have lapses in care compared with White patients who lived in a neighborhood with less deprivation. Yet, lapses in diabetic care for Black patients and Hispanic patients were consistent regardless of neighborhood deprivation, at probabilities between those of White patients in the least deprived and most deprived neighborhoods. Nuance is revealed when we explore potential patient-level and neighborhood-level interactions.

Understanding these interactions can lead to the development of more comprehensive and effective eye care and vision health interventions, but oftentimes, data are not available at both the individual-level and at the neighborhood-level. This is because research data are often stripped of potential identifiers, such as address. Ong and colleagues data are robust in that the study presents both neighborhood and patient-level data.¹ To overcome barriers related to potential identifiers in data collection, specifically address data, researchers could generate neighborhood-level measures from address data and then remove the address data from patient records. This would result in stored neighborhood data being aggregated at a geographic level, such as census block, census tract, zip code, or county. Researchers could also convert addresses to Federal Information Processing Standards codes and maintain a secured dataset to link to neighborhood level resource metrics. While geographic data from large national surveys, such as the American Community Survey, can be valuable for understanding an individual’s neighborhood, neighborhood data can also be provided directly by the individual. For example, individuals may be able to provide more detailed information on their perspectives on key neighborhood resources that may be sufficient or inadequate through qualitative semi-structured interviews.⁷ Collecting place-based information such as addresses or individual’s personal neighborhood experiences in order to better understand the neighborhood’s influence on health outcomes is an important area in the design of future eye and vision health equity research.