Evaluating Discrepancies in Self-Reported Glaucoma and Electronic Health Records in the National Institutes of Health All of Us Database.

PURPOSE

Patient understanding of glaucoma diagnosis is critical for disease management and it is unclear if there are racial/ethnic differences in this regard. The purpose of this study was to determine the level of agreement between glaucoma diagnosis by self-report and electronic health record (EHR) data using the National Institute of Health's "All of Us (AoU)" database and to examine the association between race/ethnicity and discordance of glaucoma diagnosis between self-report and EHR data.

DESIGN

Cross-sectional study.

PARTICIPANTS

Individuals in AoU who answered a survey question about glaucoma diagnosis and had EHR data availability.

METHODS

The agreement between self-reported glaucoma and EHR data was estimated using Cohen's κ coefficient. Multivariable logistic regression was performed, adjusting for age, sex, education level, income, and health care literacy, to examine the associations between race/ethnicity and discordance between self-reported glaucoma and EHR diagnosis.

MAIN OUTCOME MEASURES

Agreement between self-reported glaucoma and EHR diagnosis.

RESULTS

Of all 87 782 individuals, 1985 (2.26%) had both self-reported and EHR glaucoma, 81 781 (92.16%) had no glaucoma, 2022 (2.31%) individuals had EHR-only glaucoma, and 1994 (2.27%) had self-report-only glaucoma (Cohen's κ = 0.47). In the multivariable regression, Black or African American (adjusted odds ratio [aOR], 1.67; 95% confidence interval [CI], 1.40-1.98), Asian (aOR, 2.63; 95% CI, 1.97-3.44), and Hispanic or Latino (aOR, 1.63; 95% CI, 1.33-1.99) individuals were more likely to have EHR-only glaucoma than White individuals. Additionally, Black or African American (aOR, 2.30; 95% CI, 1.97-2.67) and Hispanic or Latino individuals (aOR,1.47; 95% CI, 1.21-1.79) were more likely to have self-report-only glaucoma compared with White individuals.

CONCLUSIONS

In the AoU database, we found that Black or African American and Hispanic or Latino individuals had higher odds of discordance between glaucoma diagnosis by self-report and EHR. Future studies are needed to examine the issues leading to this discordance, such as a lack of patient understanding regarding their diagnosis or a lack of culturally appropriate physician explanation/teaching. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.