Abstract
Purpose :
Roughly 4.2 million Americans aged 40 and older have uncorrectable vision impairment, of whom 1.02 million are legally blind. Another estimated 8 million have vision impairment due to uncorrected refractive error. Properly accommodating vision impairment in the medical setting is important for patient care, outcomes, and satisfaction, but there is not a designated field in the electronic health record (EHR) to record vision impairment status. We investigate 3 different methods for identifying these patients in the EHR and compare their effectiveness at identifying patients with vision impairment.
Methods :
We used 3 methods for identifying patients with vision impairment: 1) moderate – severe visual acuity impairment (>= 0.6 LogMAR in corrected best eye at distance), 2) Diagnosis code of low vision (ICD10 H54.* excluding codes for single eye impairment) in the problem list, and 3) answered Yes to “Are you blind or do you have serious difficulty seeing, even when wearing glasses?” on Race Ethnicity Language and Disability (REALD) questionnaire.We pulled visual acuity data, problem list diagnoses, and REALD data from the EHR (Epic) for all outpatient visits to Casey Eye Institute at OHSU during 2020 – 2022 and included only patients with complete data.
Results :
There were 90,953 patients seen during the study period and 40,806 patients with complete data. Of these, there were 5459 (13%) patients who were identified with vision impairment by at least one method; 2679 (7%) by visual acuity, 728 (2%) by a low vision diagnosis code in their problem list, and 3425 (8%) who self-identified as visually impaired on REALD. Only 179 (0.4%) patients were identified by all 3 methods, 1015 (3%) were identified by 2 methods, and 4265 (10%) were identified by 1 method. Figure 1 shows the overlap of the methods in identifying low vision patients.
Conclusions :
There is currently not a single reliable method for identifying patients with vision impairment in the EHR. Our study shows that the 3 different methods had very little overlap and low vision status may easily be missed during clinical care. In addition, clinical markers for low vision may not match patients’ perceptions of their disability. In order to accommodate these patients properly, there should be a designated field easily recognizable by providers and that includes patient input for identifying vision impairment.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.