PROJECT SUMMARY/ABSTRACT
Adverse drug reactions related to inadequately dosed or prescribed medications are a major cause of morbidity,
mortality, and hospital readmissions in the United States. Pharmacogenetics research has identified genetic variants
associated with response or toxicity for hundreds of drugs, and variant genotypes carried by some individuals may
contribute disproportionately to medication-related hospital admissions. Yet the clinical utility of pharmacogenetics
(PGx) when implemented in healthcare systems has not been widely established. In a resource-limited health system, it
is not practical to genotype every patient, but if we can identify patients at higher risk of hospital admission based on
medication prescribing events, we could be better able to discern which patients to approach with preemptive
genotyping. We propose a cohort study within two learning healthcare systems with established PGx implementation
programs – NorthShore University HealthSystem and The University of Chicago Center – in a pooled, ethnically diverse
patient population of 5,726 patients – to evaluate the aggregate impact of pharmacogenetic drug interactions on
readmissions and Emergency Department (ED) visits attributable to adverse drug reactions including inadequate drug
response, over a two-year follow-up period. The overall goal of this project is to develop novel analytic methods to
quantify the contribution of gene-x-drug interactions after hospital discharge to ED visits and readmissions, and to
identify those individuals at greatest risk of these outcomes. Aim 1 is to estimate the rate of combined 90-day ED visits
and readmissions attributable to inadequate drug response or toxicity and identify the clinical and demographic
predictors of those admissions. Aim 2 is to identify 90-day ED visits and readmissions associated with adverse drug
reactions among patients discharged on medications for which there is evidence of genetic variability for outcomes –
focusing on medications with high-level evidence guidelines from the Clinical Pharmacogenomics Implementation
Consortium and Food and Drug Administration labels prescribed and continued at hospital discharge on 90-day
readmissions. Aim 3 will be to determine the contribution of specific genotypes to medication-related 90-day ED visits
and readmissions visits among genotyped patients. This project is responsive to RFA HG-20-037 Advancing Genomic
Medicine Research by facilitating analysis of clinical genomic data and will leverage the extensive experience of our team
and dataset from two diverse, real-world healthcare systems across Chicagoland. Our multidisciplinary, collaborative
team, aided by use of natural language processing, will screen electronic health records to extract de-identified patient
data, curate, and apply innovative, integrated analytic methods to better identify patients who could benefit the most
from preemptive PGx genotyping and pharmacovigilance to prevent ED visits and readmissions. If, as we hypothesize,
our analyses demonstrate that gene-x-drug interactions contribute substantially to 90-day ED visits and readmissions,
these methods can be automated and will be transportable to other health systems to guide preemptive genotyping
protocols for the highest risk patients to prevent hospital readmissions. These results will inform the design of a future
clinical effectiveness trial of tailored, PGx interventions to prevent adverse drug reactions and associated readmissions.
