A Geodemographic Service Coverage Analysis of Travel Time to Glaucoma Specialists in Florida.

PRECIS
Approximately 1 in 9 Florida residents over the age of 65 years (438,642 people) live more than an hour away from a glaucoma specialist, which represents a significant barrier to care.


PURPOSE
To describe access to glaucoma care for Florida's elderly population using travel time to American Glaucoma Society (AGS) member offices.


METHODS
For this cross-sectional service coverage analysis, a list of AGS member locations was extracted from the AGS website, and addresses were geocoded using ArcGIS Online. Driving time regions were created using the service area tool in ArcGIS Pro 2.4 and overlaid with 2010 United States Census and 2016 American Community Survey data for all Florida residents age 65 or older.


RESULTS
Fifty-eight AGS member providers with 65 locations were identified and geocoded. There were 3,797,625 individuals aged over 65 years in Florida, of which 1,153,320 (30.4%) lived within 15 minutes of driving time from an AGS provider's office, 2,586,825 (68.1%) within 30 minutes, 3,358,983 (88.4%) within 60 minutes, and 3,491,815 (91.9%) within 120 minutes. The areas with the lowest access include rural areas near Lake Okeechobee and the Florida Panhandle. The population living beyond a 60-minute drive was more likely to be White, non-Hispanic and older, but less likely to be living below the federal poverty level or receiving public assistance than the population living within a 60-minute drive.


CONCLUSIONS
There is a significant travel burden for the elderly community of Florida to reach AGS providers. Additional studies could help identify other social barriers to accessing glaucoma providers in Florida and beyond in an effort to improve patient compliance and, ultimately, vision outcomes.

P atient compliance with follow-up is integral to optimizing management outcomes for glaucoma treatment. 1 Travel time to and from a glaucoma specialist's office is a major factor in compliance. 2 This is particularly important for patients with more severe glaucoma, as they require more frequent assessment but are less likely to travel from home 3 or be able to safely drive to their ophthalmologist. 4 Subsequently, patients with worse follow-up adherence are more likely to have severe glaucomatous damage. 5 Although many factors play a role in noncompliance, travel time is an important social determinant that merits attention. 6 Service coverage analyses rely on geospatial coding to calculate the mean travel time to specified locations, which can then be superimposed with demographic data. Previous studies have assessed the travel time to ophthalmologists as a measure of burden, 7 particularly to compare access between ophthalmologists and optometrists. 8 We found no systematic analyses of statewide travel time as a marker of glaucoma care access. In this pilot study, we performed a service coverage analysis of American Glaucoma Society (AGS) members' office addresses and analyzed the proportion of Florida residents aged 65 years or older with excessive travel times to these providers.

METHODS
No institutional review board approval was required for this study as all analyses included publicly available and de-identified data. A list of AGS members and their addresses was extracted from the "Find an AGS Doctor" webpage (https://secure. americanglaucomasociety.net/AGS/Find-An-AGS-Doctor) as listed from July 31, 2019 through August 13, 2019. If more than one address was available for a provider, each location was listed as a separate office for that provider and retained in the analysis.
We first used the cloud-based geocoding tool in ArcGIS Online (Esri, Redlands, CA) to geocode the locations of all AGS service providers using the addresses in the organization's member directory. We then generated driving time regions using the service area tool in ArcGIS Pro 2.4 (Esri), which uses Esri's cloud-based road network layer. We created regions representing 15-, 30-, 60-, and 120-minute drive times to each provider using average traffic conditions for 12:00 PM on Wednesdays. AGS doctors within Florida and those with driving time regions within 120 minutes of the Florida border were included. We used the 60-minute drive time as the threshold for low access, consistent with the Health Resources and Services Administration 60-minute drive time threshold used to score primary care health provider shortage areas. 9 We overlaid 2016 American Community Survey data at the census tract scale, to calculate differences in available social determinants of health within and beyond the AGS service areas. The proportion of each population living within or outside the 60-minute drive to AGS provider with each social determinant was compared by χ 2 test.

RESULTS
A total of 58 AGS member providers with 65 locations were identified and geocoded. There were 3,797,625 individuals aged over 65 years in Florida ( Fig. 1) (Fig. 2). There were 438,642 (11.6%) Florida residents aged over 65 years that were > 60 minutes away from the nearest AGS provider, and 305,810 (8.9%) that are > 120 minutes away. The regions with the lowest access include rural areas around Lake Okeechobee and throughout the Florida panhandle (excluding Tallahassee). Table 1 reports differences in reported social determinants of health of the populations that live within and beyond a 60-minute drive of an AGS provider in Florida. The population that lives beyond a 60-minute drive of an AGS provider had a greater proportion of individuals who were White, non-Hispanic race and age 65 years or older, and a lower percentage of households on public assistance or individuals below the federal poverty level (all P < 0.00001).

DISCUSSION
Access to glaucoma specialists is an important public health issue for millions of Americans. 10 We presented a service coverage analysis of mean travel time to AGS ophthalmologists in Florida. Approximately 1 in 9 Florida residents aged over 65 years live beyond a 60-minute drive of an AGS member provider's office, which constitutes a significant travel barrier to accessing a glaucoma specialist and may have negative consequences for their health. Interestingly, those who live beyond a 60-minute drive were more likely to be White, non-Hispanic, and older, but less likely to be living below the federal poverty level or receiving public assistance. This suggests that populations with longer travel time to AGS providers may be demographically homogenous but have more sufficient financial means for either traditional or alternative forms of glaucoma care such as telemedicine.
Previous studies have utilized service coverage analyses in ophthalmology but not at the subspecialty level. Lee  Florida. Similarly, a greater proportion of elderly (92.6% to 98.6%) lived within an hour of an ophthalmologist in these states compared with 88.4% of the elderly to an AGS member in Florida. This is likely because there are more ophthalmologists than AGS providers.
Travel plays an important role in compliance. Ramulu and colleagues 3 demonstrated not only that patients with glaucoma travel away from home less than glaucoma suspects, but each 5-dB decrement in the better eye mean deviation corresponded with a 6% decrease in time away from home. Patients with bilateral glaucoma are more likely to no longer drive and would have to rely on alternative transportation to visit their doctor. 3 Wright et al 11 explored publicly funded eye care utilization in Northern Ireland and describe the importance of car ownership in eye care utilization. They found not only substantially lower uptake of eye examinations for those without cars, but the greatest decrease in uptake for non-car owners when driving times exceeded 4 minutes, or~1.5 miles. Although the Northern Ireland population is more concentrated than that of Florida, car ownership is still likely an important factor in the current analysis, especially given that the public transportation system in Florida is less robust than other areas of the United States. Funk et al 6 recently published a study that assessed the odds of missed follow-up after trabeculectomy or glaucoma drainage device based on straight-line distance from home address to their eye center. They not only found that patients who lived > 50 miles straight distance from the clinic had greater odds of missing follow-up but also that straight-line distance to interstate highway access was an independent contributor to being lost to follow-up in their logistic model. Their work supports the importance of distance to provider but differs from the current study by calculating distance by straight-line rather than driving time.
Florida is a unique state worth consideration due to its large elderly population as well as its diverse regions and population. Because health insurance is not portable across state lines and Florida is a peninsula, the majority of the  population obtains its care in-state, further lending its feasibility as a pilot service coverage analysis with minimal geographic confounders. In the current study, the areas with lowest access included rural areas around Lake Okeechobee and the northwest Panhandle (excluding Tallahassee). In particular, the area surrounding Lake Okeechobee hosts many migrant agricultural workers, a population that is not only vulnerable to general health care disparities, but also specifically eye morbidity. 12,13 Outcomes in rural areas generally relate to infrastructure issues such as availability and access. 14 Even after controlling for poverty and health care supply, individuals in rural areas are less likely to access health care than their urban counterparts. 15 Like other subspecialties, the concentration of ophthalmologists directly correlates with population. 16 The dearth of physicians in rural areas is likely due to personal preferences in geographic location to live 17 but not retention once recruited. 18 Future studies specific to ophthalmology practice in the rural setting may provide further insight regarding eye care disparities in these vulnerable areas.
The role of race in this study is complex. A greater proportion of the elderly who lived further from a glaucoma specialist were White, non-Hispanic, reflecting the racial composition of rural Florida, which is consistent with the literature. Stein et al 8 similarly considered travel time to ophthalmologist by race and, while no statistical comparison is reported, their data similarly suggest a greater travel distance for White, non-Hispanic versus Black, non-Hispanic and Hispanic individuals. The 2001 National Household Travel Survey suggests rural individuals, of whom > 80% are White, non-Hispanic, have a greater travel burden to access medical and dental care than those in more diverse urban areas. 19 While the demographics of those further from AGS providers in Florida suggest that, as a population, their racial composition has a more favorable risk of glaucoma than those who live closer, these individuals are also older which would provide a counter risk for glaucoma. Likewise, while the racially diverse urban population may live closer to glaucoma specialists, other important impediments to care, such as economic and cultural barriers, should also be considered. Murakami et al 20 identified Black race and Latino ethnicity among factors as associated with inconsistent follow-up. Areas in Florida closer to an AGS provider have a greater proportion of these minorities at risk for inconsistent follow-up. Not only is prevalence of open-angle glaucoma in Hispanics greater than that reported for White, non-Hispanics and more comparable to African Americans 21 but they also face unique challenges in glaucoma diagnosis, treatment, and management. 22,23 In addition, South Florida is home to many people of Afro-Caribbean ancestry, a group known to have a relatively high incidence of open-angle glaucoma. 24,25 Individuals living closer to an AGS provider in Florida are more likely under the federal poverty line and receive public assistance as these economically disadvantaged individuals are more concentrated in urban centers. They may rely on public transportation system and a lack of car ownership can influence eye care utilization. 11 Although these higher risk minorities may be more likely to live closer than White, non-Hispanic Floridians to AGS providers, they also deserve attention as a population vulnerable to glaucoma.
There are several limitations to the current study. Individuals who did not participate in the census, including undocumented immigrants who may work in rural settings, could not be included. Individuals may seek glaucoma care across state lines, such as in Dothan, Georgia, or Mobile, Alabama. The driving time service areas computed for this analysis identify how close a neighborhood is to the nearest provider (ie, within 15 min), but do not account for the number of providers at a given location or service capacity. Providers were identified from the AGS website directory which may not be up-to-date. Some providers may choose not to publicly list their location or may not have listed all locations if they practice at several offices. The volume of services provided by each AGS member is not known but would provide another important factor when considering access to care. Not all ophthalmologists that provide full glaucoma management are AGS members. Comprehensive ophthalmologists, particularly in rural communities, may provide glaucoma services. In addition, some patients have their glaucoma managed medically by optometrists. However, patients with more severe, vision threatening disease would likely benefit from care by an AGS doctor. Mean travel times also did not consider public transportation schedules. We did not explore any potential relationship between travel time and clinical outcomes.
Despite these limitations, the current study demonstrates a significant travel burden to AGS members for the elderly community of Florida, which could have negative consequences for patient care. This service coverage analysis may serve as a pilot study that could be extended to the entire country or other geopolitical entities. Additional studies could help identify barriers to accessing glaucoma providers in an effort to improve patient compliance and, ultimately, vision outcomes. Public policy stakeholders could also consider these data to enhance patient access through strategies ranging from targeted provider reimbursement incentivization to fostering the development of mobile clinics or telemedicine.