The U.S. system for COVID-19 data collection has faced widespread criticism, with experts saying the system is too disconnected and reactive to produce accurate results. Below, four hospital IT execs share how they think the process can be improved.
Editor’s note: Responses have been edited lightly for clarity and style.
Bruce Darrow, MD, PhD. Deputy CIO and Chief Medical Information Officer at Mount Sinai Health System (New York City). Nationally, the integrity of data for COVID-19 infection and vaccination rates has suffered from the absence of a national patient identifier. We are cobbling together information from multiple different systems without a common standard to manage patient identity across them. To my knowledge, no other industry attempts to manage such complicated information — and at such high stakes — without the basic ability to verify the identity of their customers.
Oscar Marroquin, MD. Chief Healthcare Data and Analytics Officer at UPMC (Pittsburgh). In an ideal scenario, having a standardized way of collecting all data related to COVID-19 — from testing to hospitalizations and outcomes to vaccinations — across the nation would have allowed us to track things more systematically. The limited interoperability of healthcare data that we have in this country has proven to be an Achilles’ heel in our ability to track things systematically.
Given this, and because the U.S. had some time to prepare for the start of the pandemic, there should have been a more deliberate effort to organize the collection of a basic set of data points (just as we currently do for infection control and other issues) that federal agencies like the CDC could use to systematically track these important metrics. We should learn from these failings and set up the right infrastructure and processes to deal with future events like this one.
Within our large healthcare system, we were able to overcome some of the interoperability issues by harmonizing all our data from disparate EMRs in our Clinical Data Warehouse. That has allowed UPMC to keep track of the activity across our 40-plus-hospital system, regardless of the EMR that is used for documentation.
Donna Roach. CIO of University of Utah Health (Salt Lake City). At the University of Utah Health, we have an excellent track record of data transparency and collaboration. Our enterprise data warehouse team has been in existence for over 20 years, so we have learned how best to glean data from our systems and provide actionable results.
The main barrier we ran into was a lack of centralization when reporting to the various county health departments. In the future, a more centralized reporting repository is needed and the flow of data should be the same whether it’s coming from the hospital to the state or the state to the hospital — this will benefit the overall community.
Scott MacLean. Senior Vice President and CIO at MedStar Health (Columbia, Md.). In early 2020, we heard about a novel virus and knew very little other than it was spreading rapidly. At first, we were sending samples to the CDC to get test results, so data collection practices have evolved over time.
That said, I’ve seen robust data collection locally and other examples of aggregation for the country since the beginning. Once testing and eventually vaccinations became widely available, our health information exchanges benefited greatly. For instance, tests administered at pharmacies ended up in patient EHRs through our state health information exchange. In the future, we could expand on this interoperability for something like proof of vaccination, which could be obtained through the CLEAR app.
As far as other potential improvements, while it is challenging to manage a constantly evolving public health emergency, we can always look for better ways to communicate with our partners in the federal government to improve clarity around mandates. Finally, we could also improve our data collection by figuring out a way to count exposures that were never tested or hospitalized. Having that data may show us that the death rate is lower and the move toward herd immunity is further along.