Factors associated with COVID-19 infection and severity in a large insured patient population in metropolitan detroit

ABSTRACT

Introduction: The novel coronavirus disease 2019 (COVID-19) pandemic is a public health emergency that disparately affects older individuals and individuals with chronic health conditions. However, the relative importance of these factors in contributing to infection-related hospitalization or death is not known. Such information could help in identify groups most in need of sheltering or early vaccination. Methods: We investigated the relationship between pre-existing medical conditions and treatments with COVID-19 diagnosis, hospitalization, intensive care unit (ICU) admission, and death. We identified 48,370 individuals aged ≥20 years with detailed longitudinal clinical and pharmaceutical records by virtue of their receiving care from a single, integrated provider serving southeast Michigan and the Detroit metropolitan area. Within this group, there were 1,196 laboratory confirmed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections between March 12, 2020 and October 27, 2020. Logistic regression was used to assess the relationship between each of the individual characteristics, pre-existing diagnostic categories and the COVID-19 related outcomes. Results: Age (adjusted odds ratio [aOR] 1.27 per decade;P=2.03x10-6), African American and Asian race-ethnicity (aOR 3.00;P=2.53x10-9 and aOR 2.79;P=0.006, respectively vs. non-Hispanic white), body mass index (aOR 1.04 per unit increase;P=2.44x10-4), renal function (aOR 1.37 per mg/dL increase in serum creatinine;P=8.96x10-8), healthcare occupation (aOR 2.24, P=0.016), and household income (aOR 0.94 per $10,000 increase, P=0.032) were significantly associated with the combined outcome of COVID-19 related hospitalization or death. Many preexisting conditions were associated with COVID-19 hospitalization or death;however, any diabetes (aOR=1.98;P=1.76x10-6) and type-2 diabetes (aOR 1.95;P=3.52x10-6) were the most significant. Other notable associations included pain (aOR 1.71;P=1.07x10-4), history of supplemental oxygen or ventilation (aOR 7.41;P=4.07x10-4), living assistance (aOR 4.46;P=4.07x10-4), prior pulmonary embolism or DVT (aOR 2.61;P=0.001), and opioid use (aOR 3.99;P=0.002). Relationships between COVID-19 severity and both thrombosis and pain were additionally supported by analyzing pharmacy records of anticoagulant and opioid use. Conclusion: This large patient population study supports many of the risk factors previously associated with COVID-19 diagnosis and complications, including age, non-white race-ethnicity, and healthcare occupation. Higher body mass index, poorer renal function, and diabetes appear to be leading independent risk factors. The heightened risk observed for opioid users is novel and requires further investigation.