Abstract
Purpose :
This study aims to leverage Electronic Health Records (EHR) and Social Determinants of Health (SDOH) to analyze the risk factors associated with visual impairment (VI) in Southwest Virginia (SW VA), a historically underserved region. By integrating diverse data sources, we aim to validate and enrich self-reported VI data, contributing to a more nuanced understanding of VI risk factors and aiding in the development of a more robust, data-driven approach for vision health surveillance.
Methods :
We conducted a retrospective observational analysis of 2,351 patient records from SW VA in the UVA Health EHR database receiving eye care in the year 2022. This analysis was supplemented with SDOH variables, including age, sex, race, tobacco use, employment status, insurance status, Distressed Communities Index (DCI) score, and county median income. VI was defined as best documented visual acuity of ≤20/40 from 10 exams recorded in the EHR. Employing logistic regression models, we investigated the effects of SDOH on the risk of VI.
Results :
In our analysis, 17.4% of patients (n=409) experienced VI. Non-working patients were over two times more likely to have VI (p<0.001, 95% CI: 1.56–3.17). Age over 80 (p<0.001, OR > 1.115) and income below $57,000 (p=0.001, OR: 1.2–1.4) were significant predictors of VI. Additionally, DCI was protective against VI between 45-65 (p=0.003, OR: 0.97–0.99), while values over 65 increased VI risk (OR: 1.00–1.82). Higher income (> $65,000) had a protective effect (OR < 0.75).
Conclusions :
Integrating EHR and SDOH data is a potent tool for identifying patient and community-level risk factors for VI. In our study, VI prevalence was high at 17.4%, surpassing the national rate of 6.98% from the Intelligent Research in Sight EHR registry. This elevated prevalence in our SW VA patient population may stem from local barriers like delayed care-seeking and the use of academic referral practice data biased toward higher care complexity. Our findings highlight the impact of SDOHs on VI and vision health disparities, providing foundational insights for leveraging integrated data approaches to enhance vision health surveillance at state and national levels.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.