EHR-based Genomic Discovery and Implementation

PROJECT SUMMARY In this application we propose to build on our work in the previous three phases of eMERGE to develop and validate polygenic risk scores (PRS) for 15 common diseases and combine these with clinical risk scores and family history to create comprehensive risk profiles for 2500 individuals (35% non-European ancestry). We will develop phenotyping algorithms to identify cases and controls for the 15 diseases and ascertain clinical risk scores for each disease by extracting relevant variables from the electronic health record (EHR). Next, using existing eMERGE genotype and phenotype data (n=105,000) and publicly available (TopMed, AoU, MVP, UKBB) datasets, we will calculate candidate loci or genome wide PRS using novel methods to adapt these to specific racial/ethnic groups. Validated PRS will be combined with family history and clinical risk scores to create comprehensive risk profiles. We have previously applied such methodology for coronary heart disease, positioning us well to expand this to 14 additional common diseases. To assess the utility of PRS in a prospective cohort, we will recruit participants from Rochester MN (1800 adults and 200 adolescents) and Phoenix AZ (500 adults of Hispanic ethnicity). DNA will be sent for CLIA-certified genomic testing and we will calculate race/ethnicity specific PRS. Individualized risk estimates will be reported to participants and providers and placed in the EHR with linkage to clinical decision support (CDS) that includes guideline-based risk- management information. For participants who are at significantly higher risk for developing at least 1 of the 15 diseases (estimated 25% of the cohort), results will be disclosed in-person followed by assessment of outcomes including new tests ordered, risk reduction measures, and changes in modifiable risk factors. In addition we will examine the economic, behavioral and psychosocial implications of returning PRS. Our focus on PRS brings genomic medicine to the population and we will continue our partnership with Mountain Park Health Center, a FQHC serving low-income, minority communities across Phoenix AZ, to address concerns about health disparities in the context of genomic medicine. Our specific aims include: 1) Utilize extant genotype and EHR data on 105,000 eMERGE participants as well as other publicly available datasets to validate PRS for 15 common diseases; 2) Deploy novel methods for integrating PRS in the EHR with linkage to CDS that incorporates guideline based clinical recommendations and develop scalable methods for disclosing PRS results to participants; 3) Enroll 2500 persons to undergo CLIA genotyping using dense SNP arrays and return comprehensive risk profiles that include PRS; 4) Assess clinical, cost/utilization and psychosocial outcomes following return of results.