Coronavirus Disease 2019 (COVID-19) caused by SARS-CoV-2 is a currently a global pandemic. While most
COVID-19 cases are mild, severe cases (~15%) require hospitalization, critical cases (~5%) require intensive
care, and many deaths occur. Males and blacks are at greater risk for COVID-19 infection, while poor prognosis
is predicted by older age, race/ethnicity, and prior underlying medical conditions. A potentially important
complicating risk factor is substance use or substance use disorders (SU/SUD). SU/SUD could increase the risk
for COVID-19 infection and poor prognosis by direct effects of the substances on the cardiovascular, respiratory
or immune systems, or by indirect effects due to the greater prevalence of underlying medical conditions among
substance users that predict poor COVID-19 prognosis. Little is known about the relationship between SU/SUD
and the risk for COVID-19 infection or poor prognosis, and whether these relationships are modified by
demographic characteristics (sex, age, race/ethnicity, homelessness), medical conditions (e.g., cardiovascular,
respiratory conditions, HIV) or state policies (marijuana laws; social distancing policies). To study these
relationships, large databases must include SU/SUD, demographic characteristics, diagnostic, treatment and
mortality information.
Responding to PA-18-935 and NOT-DA-20-047, we will utilize the Veterans Administration
(VA) Electronic Medical Record (EMR) system for this purpose. The VA treats 5.7 million veterans a year. VA
patients have high rates of COVID-19
vulnerability factors, e.g., male, older age, and chronic medical conditions.
A VA Shared Data Resource identifies COVID-19 cases. The large number of VA patients with ICD-10-CM SUD
diagnoses or positive substance use screens will provide extensive data on whether the risk for COVID-19
infection and poor prognosis differs by SU/SUD status. Leveraging the research infrastructure established in our
parent grant R01DA048860, we propose a 2-year Competitive Revision to comprehensively address the
relationship of SU/SUD to COVID-19 infection and prognosis, and if this varies by demographic, medical and
state characteristics. Aim 1: Determine if SU/SUD (cannabis, tobacco, opioid, stimulants or cocaine) increase
the risk for COVID-19 infection, or in those infected, poor prognosis (e.g., hospitalization, ICU treatment,
intubation, death). Aim 2: Determine if associations of SU/SUD with COVID-19 outcomes vary by demographic
(sex, age, race/ethnicity, homelessness), clinical (e.g., underlying cardiovascular or respiratory conditions, HIV),
or state characteristics. In Year 01, we will focus on 2020 EMR diagnostic, treatment, and vital status death data,
using 2019 data to establish that SU/SUD preceded COVID-19. In Year 02, we will incorporate Medicare data to
expand information on those =age 65, and incorporate National Death Index data to examine causes of death.
Logistic regression will evaluate differences in COVID-19 outcomes by SU/SUD status. Among those with
COVID-19, survival models will determine if time to events indicating poor prognosis differs by SU/SUD. Results
will fill a major gap in knowledge about the risks for and prognosis of COVID-19 among those with SU/SUD.