Policy makers have a critical role to play in the COVID-19 pandemic. Their decisions about how fast to reopen society and lift shelter-in-place orders will directly impact infection rates, and also determine how severely people are harmed by policies that continue to restrict the population’s ability to work and interact with other people. These are excruciating decisions. While policy makers have easy access to daily counts of cases, hospitalizations, and deaths from COVID-19, they currently do not have access to the patient perspective. Policy makers need information about how restrictive policies are actually impacting the populations they serve, and how successful their policies are at providing access to resources meant to contain and mitigate effects of the Pandemic. Our proposal will compare the impact of policy maker decisions in different states, counties, and health systems across the US, and will focus particularly on the impact in communities that are most vulnerable to harms due to their economic situation, employment, housing, location (e.g., rural or low-income) or burden of medical comorbidities. Our specific research questions are:
- What is the comparative impact of different shelter-in-place/reopening policies, overall and in vulnerable populations, on patient-reported financial insecurity, mental health, and other subjective outcomes important to patients?
- What is the comparative effectiveness of county-level containment and mitigation strategies at achieving timely access to testing, healthcare, information, and contact tracing, overall and in vulnerable populations?
- What is the comparative accuracy of different algorithms designed to predict risk of infection and severe COVID-19 among patients with symptoms, overall and in vulnerable populations?
To answer these questions, we must directly ask the people in harm’s way. To do this, we have launched COVID-19 Citizen Science, a digital cohort study that has already recruited over 20,000 individuals across the US that are self-reporting daily symptoms of COVID-19 along with COVID-19-relevant behaviors and patient-reported outcomes. However, our sample is not yet representative of the US population with underrepresentation of vulnerable populations, is currently concentrated in California, and does not include critical healthcare-derived information such as test results and hospitalizations that we need to link to our patient-reported outcomes.
For this project, we will recruit a population-based sample of Citizen Scientists from PCORnet® in seven critical US states with different politics, demographics, and COVID-19 dynamics, with heavy emphasis on enrollment of vulnerable populations and reduction of the selection biases that otherwise naturally occur. Our patient-generated data will be linked to electronic health record data and claims data from PCORnet using methods we have established and currently use in our PCORnet Blood Pressure Control Laboratory. Our study design, data collection, recruitment, and dissemination strategies will all be informed by deep engagement with a multi-stakeholder advisory board that includes Citizen Scientists representing vulnerable communities and policy makers from states, counties, and health systems for whom our results will be immediately actionable. Our investigator team has deep expertise in epidemiology, policy, vulnerable populations and data science, and many work together already in large PCORnet-based consortia.