Extracting Primary Open-Angle Glaucoma from Electronic Medical Records for Genetic Association Studies

Abstract

Electronic medical records (EMRs) are being widely implemented for use in genetic and genomic studies. As a phenotypic rich resource, EMRs provide researchers with the opportunity to identify disease cohorts and perform genotype-phenotype association studies. The Epidemiologic Architecture for Genes Linked to Environment (EAGLE) study, as part of the Population Architecture using Genomics and Epidemiology (PAGE) I study, has genotyped more than 15,000 individuals of diverse genetic ancestry in BioVU, the Vanderbilt University Medical Center’s biorepository linked to a de-identified version of the EMR (EAGLE BioVU). Here we develop and deploy an algorithm utilizing data mining techniques to identify primary open-angle glaucoma (POAG) in African Americans from EAGLE BioVU for genetic association studies. The algorithm described here was designed using a combination of diagnostic codes, current procedural terminology billing codes, and free text searches to identify POAG status in situations where gold-standard digital photography cannot be accessed. The case algorithm identified 267 potential POAG subjects but underperformed after manual review with a positive predictive value of 51.6% and an accuracy of 76.3%. The control algorithm identified controls with a negative predictive value of 98.3%. Although the case algorithm requires more downstream manual review for use in large-scale studies, it provides a basis by which to extract a specific clinical subtype of glaucoma from EMRs in the absence of digital photographs.

Local genetic ancestry in CDKN2B-AS1 is associated with primary open-angle glaucoma in an African American cohort extracted from de-identified electronic health records

“…A total of 138 African American POAG cases and 1376 controls passed quality control in EAGLE BioVU for the present study. We previously described [ 34 ] these cases compared with 4813 controls over the age of 40; in the present study, we compare the same cases with subset of the controls over the age of 60 years (Table 1 ). Here, cases were younger ( p = 0.01) with higher cholesterol levels (183 mg/dL; p = 0.01), more likely to be female with higher average body mass index (30.1 kg/m 2 ) in comparison to controls (28.8 kg/m 2 and 169 mg/dL, respectively).…”

“… Case extraction was described in Restrepo et al [ 34 ]. Control extraction from EAGLE BioVU was also described in Restrepo et al [ 34 ] but restricted to controls > 40 years of age (as opposed to > 60 years of age here).…”

“…From among these patients, billing and procedural codes along with text searches were used to identify POAG cases ( n = 138) and controls ( n = 1376). In short, controls included patients in BioVU over the age of 60 years whose records did not contain an ICD-9-CM code for any form of glaucoma nor any mention of “glaucoma” in a text search of their ‘Problems List.’ Manual review of all cases and a subset of controls was performed for quality assurance as previously described [ 34 ].…”

Primary Open-Angle Glaucoma Genetics in African Americans

“…No significant signals were detected in the Ghanaian samples. In another study of African Americans with POAG, Restrepo and colleagues did not find that SNPs in this gene contributed to POAG risk even after performing a fine-mapping analysis that included 286 SNPS within CDKN2B-AS1 (59). …”

“…Genetic African Ancestry was significantly associated with both thinner CCT and higher IOP in POAG patients; however, when stratified by genetic African ancestry (applying a cutoff of 60%), CCT was not associated with POAG, suggesting that thinner CCT is not associated with POAG in this African population sample. The use of African genetic ancestry as a defining variable has also shown some promise for elucidating complex genetic associations in other studies, such as in Restrepo et al ., where percentage of African Ancestry in an African American cohort was associated with POAG at the CDKN2B-AS1 locus, even though single SNP analysis failed to identify such a correlation (59). …”