A phenome-wide association study (PheWAS) of COVID-19 outcomes by race using the electronic health records data in Michigan Medicine (preprint)

Blacks/African Americans are overrepresented in the number of hospitalizations and deaths from COVID-19 in the United States, which could be explained through differences in the prevalence of existing comorbidities. We performed a disease-disease phenome-wide association study (PheWAS) using data representing 5,698 COVID-19 patients from a large academic medical center, stratified by race. We explore the association of 1,043 pre-occurring conditions with several COVID-19 outcomes: testing positive, hospitalization, ICU admission, and mortality. Obesity, iron deficiency anemia and type II diabetes were associated with susceptibility in the full cohort, while ill-defined descriptions/complications of heart disease and stage III chronic kidney disease were associated among non-Hispanic White (NHW) and non-Hispanic Black/African American (NHAA) patients, respectively. The top phenotype hits in the full, NHW, and NHAA cohorts for hospitalization were acute renal failure, hypertension, and insufficiency/arrest respiratory failure, respectively. Suggestive relationships between respiratory issues and COVID-19-related ICU admission and mortality were observed, while circulatory system diseases showed stronger association in NHAA patients. We were able to replicate some known comorbidities related to COVID-19 outcomes while discovering potentially unknown associations, such as endocrine/metabolic conditions related to hospitalization and mental disorders related to mortality, for future validation. We provide interactive PheWAS visualization for broader exploration.

COVID-19 outcomes, risk factors and associations by race: a comprehensive analysis using electronic health records data in Michigan Medicine (preprint)

Structured AbstractO_ST_ABSImportanceC_ST_ABSBlacks/African-Americans are overrepresented in the number of COVID-19 infections, hospitalizations and deaths. Reasons for this disparity have not been well-characterized but may be due to underlying comorbidities or sociodemographic factors. ObjectiveTo systematically determine patient characteristics associated with racial/ethnic disparities in COVID-19 outcomes. DesignA retrospective cohort study with comparative control groups. SettingPatients tested for COVID-19 at University of Michigan Medicine from March 10, 2020 to April 22, 2020. Participants5,698 tested patients and two sets of comparison groups who were not tested for COVID-19: randomly selected unmatched controls (n = 7,211) and frequency-matched controls by race, age, and sex (n = 13,351). Main Outcomes and MeasuresWe identified factors associated with testing and testing positive for COVID-19, being hospitalized, requiring intensive care unit (ICU) admission, and mortality (in/out-patient during the time frame). Factors included race/ethnicity, age, smoking, alcohol consumption, healthcare utilization, and residential-level socioeconomic characteristics (SES; i.e., education, unemployment, population density, and poverty rate). Medical comorbidities were defined from the International Classification of Diseases (ICD) codes, and were aggregated into a comorbidity score. ResultsOf 5,698 patients, (median age, 47 years; 38% male; mean BMI, 30.1), the majority were non-Hispanic Whites (NHW, 59.2%) and non-Hispanic Black/African-Americans (NHAA, 17.2%). Among 1,119 diagnosed, there were 41.2% NHW and 37.4% NHAA; 44.8% hospitalized, 20.6% admitted to ICU, and 3.8% died. Adjusting for age, sex, and SES, NHAA were 1.66 times more likely to be hospitalized (95% CI, 1.09-2.52; P=.02), 1.52 times more likely to enter ICU (95% CI, 0.92-2.52; P=.10). In addition to older age, male sex and obesity, high population density neighborhood (OR, 1.27 associated with one SD change [95% CI, 1.20-1.76]; P=.02) was associated with hospitalization. Pre-existing kidney disease led to 2.55 times higher risk of hospitalization (95% CI, 1.62-4.02; P