Project Summary. As of April 30, 2020, over 1 million individuals in the U.S. have been diagnosed with
coronavirus disease 2019 (COVID-19). Patients with COVID-19 may develop various symptoms – while the
majority of patients have mild symptoms, some require hospitalization, admissions to intensive care unit (ICU),
and may die. To date, there is only limited knowledge on risk factors associated with the severity of COVID-19.
First, older adults have been found to have higher risks of developing severe symptoms of COVID-19 and are
more likely to be hospitalized or die. Studies have suggested that some underlying conditions, such as
hypertension, diabetes, or obesity, are associated with the severity of COVID-19. However, it is unknown to
what extent these comorbidities explain the variation in the severity of COVID-19, whether older age is
independently associated with the severity of COVID-19; and whether and how older age modifies the
relationship between comorbidities and the severity of COVID-19. Second, it has been reported that black
Americans experienced a higher rate of COVID-related hospitalization and were more likely to die of COVID-
19, compared to white Americans. However, it is unknown what may contribute to such racial difference –
whether it is due to the differences in health conditions between blacks and whites, or due to the
characteristics of the community where they reside in, or due to some other factors that are also associated
with race. The objective of this study is to identify individual risk factors that are associated with the severity of
COVID-19 (i.e. hospitalizations, ICU use and death), especially among older adults, and to understand reasons
that may contribute to racial differences in COVID-19 severity. To achieve these goals, we will use the daily-
updated national Veterans Affairs (VA) data, which contain rich individual-level information on veterans
diagnosed with COVID-19. As of April 30, 2020, almost 9,000 veterans have been diagnosed with COVID-19,
and about 500 had died, thus providing a large study cohort. This proposed study has two Specific Aims:1) To
identify individual risk factors that are associated with COVID-19 related hospitalizations, ICU use and
mortality, to understand the role of older age in COVID-19 severity, and to build a predictive model for COVID-
19 severity by machine learning; and 2) To examine reasons for racial differences in illness severity among
veterans diagnosed with COVID-19: whether and how such difference is related to individual factors and
community characteristics, especially socio-economic status. This study is innovative because it will be the first
study to examine the role of multiple risk factors in the severity of COVID-19 by using national data with
detailed individual-level information and machine learning algorithm; and it will be the first to examine the
reasons, including the role of social determinants, for racial differences in COVID-19 severity. This proposed
research is significant as it will help to identify patients with the highest-risk phenotypes, thus providing insights
into disease prevention and resource allocation.