Motivation: Emergence of genetic data coupled to longitudinal electronic medical records (EMRs) offers the possibility of phenome-wide association scans (PheWAS) for disease–gene associations. We propose a novel method to scan phenomic data for genetic associations using International Classification of Disease (ICD9) billing codes, which are available in most EMR systems. We have developed a code translation table to automatically define 776 different disease populations and their controls using prevalent ICD9 codes derived from EMR data. As a proof of concept of this algorithm, we genotyped the first 6005 European–Americans accrued into BioVU, Vanderbilt's DNA biobank, at five single nucleotide polymorphisms (SNPs) with previously reported disease associations: atrial fibrillation, Crohn's disease, carotid artery stenosis, coronary artery disease, multiple sclerosis, systemic lupus erythematosus and rheumatoid arthritis. The PheWAS software generated cases and control populations across all ICD9 code groups for each of these five SNPs, and disease-SNP associations were analyzed. The primary outcome of this study was replication of seven previously known SNP–disease associations for these SNPs.Results: Four of seven known SNP–disease associations using the PheWAS algorithm were replicated with P-values between 2.8 × 10−6 and 0.011. The PheWAS algorithm also identified 19 previously unknown statistical associations between these SNPs and diseases at P < 0.01. This study indicates that PheWAS analysis is a feasible method to investigate SNP–disease associations. Further evaluation is needed to determine the validity of these associations and the appropriate statistical thresholds for clinical significance.Availability:The PheWAS software and code translation table are freely available at http://knowledgemap.mc.vanderbilt.edu/research.Contact: josh.denny@vanderbilt.edu

Candidate gene and genome-wide association studies (GWAS) have identified genetic variants that modulate risk for human disease; many of these associations require further study to replicate the results. Here we report the first large-scale application of the phenome-wide association study (PheWAS) paradigm within electronic medical records (EMRs), an unbiased approach to replication and discovery that interrogates relationships between targeted genotypes and multiple phenotypes. We scanned for associations between 3,144 single-nucleotide polymorphisms (previously implicated by GWAS as mediators of human traits) and 1,358 EMR-derived phenotypes in 13,835 individuals of European ancestry. This PheWAS replicated 66% (51/77) of sufficiently powered prior GWAS associations and revealed 63 potentially pleiotropic associations with P < 4.6 × 10−6 (false discovery rate < 0.1); the strongest of these novel associations were replicated in an independent cohort (n = 7,406). These findings validate PheWAS as a tool to allow unbiased interrogation across multiple phenotypes in EMR-based cohorts and to enhance analysis of the genomic basis of human disease.

The Electronic Medical Records and Genomics (eMERGE) Network is a National Human Genome Research Institute (NHGRI)-funded consortium engaged in the development of methods and best-practices for utilizing the Electronic Medical Record (EMR) as a tool for genomic research. Now in its sixth year, its second funding cycle and comprising nine research groups and a coordinating center, the network has played a major role in validating the concept that clinical data derived from EMRs can be used successfully for genomic research. Current work is advancing knowledge in multiple disciplines at the intersection of genomics and healthcare informatics, particularly electronic phenotyping, genome-wide association studies, genomic medicine implementation and the ethical and regulatory issues associated with genomics research and returning results to study participants. Here we describe the evolution, accomplishments, opportunities and challenges of the network since its inception as a five-group consortium focused on genotype-phenotype associations for genomic discovery to its current form as a nine-group consortium pivoting towards implementation of genomic medicine.

Despite the diverse structure of the five EMRs of the eMERGE sites, we developed, validated, and successfully deployed 13 electronic phenotype algorithms. Validation is a worthwhile process that not only measures phenotype performance but also strengthens phenotype algorithm definitions and enhances their inter-institutional sharing.

We repurposed existing genotypes in DNA biobanks across the Electronic Medical Records and Genomics network to perform a genome-wide association study for primary hypothyroidism, the most common thyroid disease. Electronic selection algorithms incorporating billing codes, laboratory values, text queries, and medication records identified 1317 cases and 5053 controls of European ancestry within five electronic medical records (EMRs); the algorithms' positive predictive values were 92.4% and 98.5% for cases and controls, respectively. Four single-nucleotide polymorphisms (SNPs) in linkage disequilibrium at 9q22 near FOXE1 were associated with hypothyroidism at genome-wide significance, the strongest being rs7850258 (odds ratio [OR] 0.74, p = 3.96 × 10(-9)). This association was replicated in a set of 263 cases and 1616 controls (OR = 0.60, p = 5.7 × 10(-6)). A phenome-wide association study (PheWAS) that was performed on this locus with 13,617 individuals and more than 200,000 patient-years of billing data identified associations with additional phenotypes: thyroiditis (OR = 0.58, p = 1.4 × 10(-5)), nodular (OR = 0.76, p = 3.1 × 10(-5)) and multinodular (OR = 0.69, p = 3.9 × 10(-5)) goiters, and thyrotoxicosis (OR = 0.76, p = 1.5 × 10(-3)), but not Graves disease (OR = 1.03, p = 0.82). Thyroid cancer, previously associated with this locus, was not significantly associated in the PheWAS (OR = 1.29, p = 0.09). The strongest association in the PheWAS was hypothyroidism (OR = 0.76, p = 2.7 × 10(-13)), which had an odds ratio that was nearly identical to that of the curated case-control population in the primary analysis, providing further validation of the PheWAS method. Our findings indicate that EMR-linked genomic data could allow discovery of genes associated with many diseases without additional genotyping cost.

By providing a central repository, PheKB enables improved development, transportability, and validity of algorithms for research-grade phenotypes using health care generated data.

Large-scale DNA databanks linked to electronic medical record (EMR) systems have been proposed as an approach for rapidly generating large, diverse cohorts for discovery and replication of genotype-phenotype associations. However, the extent to which such resources are capable of delivering on this promise is unknown. We studied whether an EMR-linked DNA biorepository can be used to detect known genotype-phenotype associations for five diseases. Twenty-one SNPs previously implicated as common variants predisposing to atrial fibrillation, Crohn disease, multiple sclerosis, rheumatoid arthritis, or type 2 diabetes were successfully genotyped in 9483 samples accrued over 4 mo into BioVU, the Vanderbilt University Medical Center DNA biobank. Previously reported odds ratios (OR(PR)) ranged from 1.14 to 2.36. For each phenotype, natural language processing techniques and billing-code queries were used to identify cases (n = 70-698) and controls (n = 808-3818) from deidentified health records. Each of the 21 tests of association yielded point estimates in the expected direction. Previous genotype-phenotype associations were replicated (p < 0.05) in 8/14 cases when the OR(PR) was > 1.25, and in 0/7 with lower OR(PR). Statistically significant associations were detected in all analyses that were adequately powered. In each of the five diseases studied, at least one previously reported association was replicated. These data demonstrate that phenotypes representing clinical diagnoses can be extracted from EMR systems, and they support the use of DNA resources coupled to EMR systems as tools for rapid generation of large data sets required for replication of associations found in research cohorts and for discovery in genome science.

IMPORTANCEThe National COVID Cohort Collaborative (N3C) is a centralized, harmonized, highgranularity electronic health record repository that is the largest, most representative COVID-19 cohort to date. This multicenter data set can support robust evidence-based development of predictive and diagnostic tools and inform clinical care and policy.OBJECTIVES To evaluate COVID-19 severity and risk factors over time and assess the use of machine learning to predict clinical severity. DESIGN, SETTING, AND PARTICIPANTSIn a retrospective cohort study of 1 926 526 US adults with SARS-CoV-2 infection (polymerase chain reaction >99% or antigen <1%) and adult patients without SARS-CoV-2 infection who served as controls from 34 medical centers nationwide between January 1, 2020, and December 7, 2020, patients were stratified using a World Health Organization COVID-19 severity scale and demographic characteristics. Differences between groups over time were evaluated using multivariable logistic regression. Random forest and XGBoost models were used to predict severe clinical course (death, discharge to hospice, invasive ventilatory support, or extracorporeal membrane oxygenation). MAIN OUTCOMES AND MEASURESPatient demographic characteristics and COVID-19 severity using the World Health Organization COVID-19 severity scale and differences between groups over time using multivariable logistic regression. RESULTSThe cohort included 174 568 adults who tested positive for SARS-CoV-2 (mean [SD] age, 44.4 [18.6] years; 53.2% female) and 1 133 848 adult controls who tested negative for SARS-CoV-2 (mean [SD] age, 49.5 [19.2] years; 57.1% female). Of the 174 568 adults with SARS-CoV-2, 32 472(18.6%) were hospitalized, and 6565 (20.2%) of those had a severe clinical course (invasive ventilatory support, extracorporeal membrane oxygenation, death, or discharge to hospice). Of the hospitalized patients, mortality was 11.6% overall and decreased from 16.4% in March to April 2020 to 8.6% in September to October 2020 (P = .002 for monthly trend). Using 64 inputs available on the first hospital day, this study predicted a severe clinical course using random forest and XGBoost models (area under the receiver operating curve = 0.87 for both) that were stable over time. The factor most strongly associated with clinical severity was pH; this result was consistent across machine learning methods. In a separate multivariable logistic regression model built for inference, (continued) Key Points Question In a US data resource large enough to adjust for multiple confounders, what risk factors are associated with COVID-19 severity and severity trajectory over time, and can machine learning models predict clinical severity? Findings In this cohort study of 174 568 adults with SARS-CoV-2, 32 472 (18.6%) were hospitalized and 6565 (20.2%) were severely ill, and first-day machine learning models accurately predicted clinical severity. Mortality was 11.6%