COVID-19 Research Database Pools Data From Millions to Support Research, Clinical Efforts, and More

by Bradley White, CEO, HHS Technology Group — Feb 15, 2023

When the pandemic began, vendors in the United States collaborated to develop the COVID-19 Research Database. The database offers real-world data to policymakers, researchers, clinicians, and other stakeholders across the country.

In this interview with Integrated Healthcare Executive, Brett Furst explains why the database was created, how the data has been leveraged for research into broad populations and specific diseases, and how health systems may use the database as the SARS-CoV-2 virus continues to evolve.

How has data in the research database been collected?

In March 2020 at the beginning of the COVID-19 pandemic, several reputable data vendors in the United States came together in a consortium to try to aggregate data in meaningful ways. The intent was to draw insight into the pandemic with scientific rigor that was not possible to do in data silos. These efforts were research-intensive, meaning there was pro bono access. We wanted our most notable epidemiologists and biostatisticians leaning into the data to help us find that insight.

The database became incredibly successful, drawing from almost 90 billion records over 300 million people. It covers claims data, clinical data, and more than 1500 different fields of social determinant data. It has now been a resource for more than 400 public publications on a range of topics, from population health and a broader look at what we can expect from long COVID-19 to very specific research like the effects of COVID-19 on dementia and Alzheimer disease.

Along that journey, that consortium approached HHS Technology Group to request we manage the database because it became so large. We have a data platform called Discover Your Data, which is an industry-hardened solution to facilitate collaborative data mining in both public and commercial sectors. With partners like Mathematica, Datavant, and others, we created this platform where people could analyze data securely while protecting the identities of the citizens we were measuring.

Our partners update the data every month. We still see signals of long COVID-19, so we are going to continue powering research moving forward. And, certainly, the research database is not just about COVID-19. It came about because of COVID-19, but it can be expanded for much broader and more generic insights, such as heart disease, diabetes, or Alzheimer disease as a whole.

Can you describe some of the most important ways the database has been used?

We try to reduce the noise surrounding data so we can see what is actually happening. This has been helpful to drug research, clinical approaches, and population studies.

Are the boosters working? Are there disparities between those who have been vaccinated vs those who have not? What can we learn about social determinants of health and health equity, such as age, race, location, and access to care?

As we supposedly wean out of the COVID-19 pandemic era, the effects of long COVID-19 are alarming. Our data show the neurological effects of long COVID-19 are persisting. People coined phrases like “the COVID fog” and whatnot, but the effect is a little more sustainable than that.

We expect cases of Alzheimer disease and dementia to triple or quadruple over the next 5 years in vulnerable populations. NIH and other private sector solutions are applying funding toward long COVID-19 because, quite frankly, Alzheimer disease and dementia will be one of our bellwether concerns for the next 5 years, based on the average age of our population.

As another example, in Michigan, we have a project related to chronic kidney disease (CKD) and COVID-19. In Michigan, about 46% of all deaths from COVID-19 were related to kidney failure. If you were a patient trending toward hypertension, diabetes, and CKD, and you had a severe case of COVID-19, we predict CKD stages are going to accelerate quicker for you than most.

CKD is called “a silent killer” because most people progress to stage 3 and do not even know they have it. We are working closely with the National Kidney Foundation in Michigan to determine how we can implement interventions, educate, and help patients with CKD.

With the research database, we are drawing that kind of insight on what we can do proactively to help populations that have underlying conditions. Early interventions are better for patients and costs of care. The database allows us to analyze whether there are issues barring access to care based on where we see signals of long COVID-19, in rural Michigan vs urban centers of Michigan, for example.

This database also has had a huge impact on state and insured budgets.

Thank you, Brett. As the virus continues evolving each season, how can people working in health systems use the database to inform decision-making in real time?

I will address this question on 2 fronts.

Across the United States, health systems typically are the largest employers in their regions. I sit on the board of the Michigan Health & Hospital Association, and I continuously hear about the pressure that COVID-19 has placed on health systems. A health system, as a self-insured employer and caregiver, may use our data on the effects of long COVID-19 on patients with underlying conditions to implement and deploy solutions to their workforce. For example, if dementia or Alzheimer disease develop earlier, these diseases make it harder for people to function normally and have a big effect on behavioral health, but patients can be treated. Health systems can leverage our data that way.

The other consideration is chronic disease management, especially as many health systems explore how they can implement better care in the home through home health models, partnering, and their own investments. A lot of helpful interventions involve lifestyle choices, coaching, education, and outreach. If health systems are not taking on risk like an accountable care organization, they must try to manage that capitated market themselves as efficiently as they can.

In our database, all this information is there for exploration, and we can zero in on a health system’s population. If a health system wants to determine where they can improve access to care, we can narrow it down by treatment options and types of cases. Are access issues related to health equity? Is it because we have an above-average age in our population due to our location? These are the issues we are trying to help with to inform health systems’ clinical decision-making strategies so they can make the proper investments for improving care in their community.

Is there anything we have not mentioned yet that you would like to add?

I would dispel the notion that we are out of the forest with COVID-19. I think there will be interesting studies on whether patients should pivot depending on which vaccine or boosters they received. We will continue to see different variants of the virus that have more resilience against what we have planned for.

On the positive side, COVID-19 brought competitors together to do something for the common good. The federal government allowed us to start merging data whereas 5 to 10 years ago, folks were not so keen on that because of fears related to privacy, antitrust, and usage. We learned a really good lesson by creating this relational database for research and public good that can persist on a broader scale.