Skip to main content

  • Do HPSAs Apply to Eye Care?

    By Lynda Seminara
    Selected by Russell N. Van Gelder, MD, PhD
    Add to My Bookmarks

    Journal Highlights

    Ophthalmology, April 2023

    Download PDF

    The U.S. government uses health professional shortage areas (HPSAs) to gauge patient accessibility to primary care, but whether these shortage areas apply to eye care is not known. Soares et al. conducted a cross-sectional retrospective study to explore this matter and found only a weak correlation between HPSAs and the availability of eye care.

    For this work, the investigators gath­ered U.S. census data and documented practice locations of ophthalmologists and optometrists listed in the Medicare database, from which eye care provider (ECP) density per county was calculat­ed. HPSA scores were determined from the publicly available HPSA dataset, and county-level data on visual impairment were derived from the CDC’s Vision and Eye Health Surveillance System. The main outcome measure was the relationship between vision loss and ECP density. Sociodemographic factors linked to ECP shortage also were explored.

    The study authors applied a two-step floating catchment area (FCA) method to newly identify areas of eye care shortage, based on patient accessibility score (PAS), with higher scores denoting lower access. For every U.S. county, they weighted the two-step FCA scores by prevalence of vision loss and ECP density.

    Altogether, there were 72,735 cen­sus tracts. The authors found weak correlations between HPSA scores and visual impairment as well as ECP density per county population. All told, 54% of census tracts with ECP density per county below the 25th percentile were HPSAs. Among the census tracts exceeding the 75th percentile for visual impairment, 58% were HPSAs (p < .0001). Multivariable regression showed higher odds of ECP PAS ≥75th per­centile (weaker accessibility) in rural counties (adjusted OR [aOR], 2.47; p < .001) and in counties with a high pro­portion of residents who lacked a high school education (aOR, 1.21; p < .001). Higher odds of ECP PAS also were more common in areas with a greater percentage of older adults (aOR, 1.10; p < .001) and people without health insurance (aOR, 1.04; p < .001). Lower odds of ECP PAS ≥75th percentile were found for counties with larger percent­ages of men (aOR, .93; p < .001) and White residents (aOR, .99; p < .001).

    The original article can be found here.