Abstract
Purpose :
Open-angle glaucomas are challenging and usually silent disease processes that often present with advanced optic nerve damage and visual field defects. In the current study, we present a novel population-based data mining analysis to gain a deeper
understanding of glaucoma risk factors through innovations in big data science. This study aims to more clearly define whether there is a correlation between neoplastic conditions and other assorted demographic variables on the development of open-angle glaucoma.
Methods :
Using deidentified electronic health records (EHR) from the vast i2b2 database from the University of Missouri, ICD (International Classification of Disease) codes based on inclusion and exclusion criteria of the specific study cohorts were examined. Cumulatively, 14,212 patient records were analyzed. We applied a Pearson correlation analysis and stepwise regression analysis to find a possible association between glaucoma and selected study variables.
Results :
This study specifically found that neoplasms of skeletal and connective tissue have an increased association (p<0.05) with open-angle glaucomas as compared to any other neoplastic processes. No other oncologic conditions appeared to show a similar association. Similarly, in our cohort, there is an association (p<0.01) between glaucoma and lifelong non-smoking status and patients with any previous history of smoking. This is particularly interesting, as patients who were current smokers were not associated with an increased glaucoma risk. Lastly, this study found that African American individuals have an incredibly statistically significant (p<0.01) risk of glaucoma development versus other study populations.
Conclusions :
In our analysis of a large subset of patients in a vast database, glaucoma was associated with neoplasms of the connective tissue and bone more than any other neoplastic process. A possible explanation for this could be the fact that there is a common ectodermal embryology between these structures. Furthermore, African Americans and nonsmokers/former smokers have been associated with an increased risk for glaucoma, when compared with current smokers, and these facts should be examined in larger scale studies. Without a doubt, this study demonstrates the vast potential of applying data mining methods for risk assessments using large-scaled EHR data.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.