3 Ways the Industry Is Leveraging Real-World Data to Combat COVID-19

Examining utilization rates, hospital capacity

The rapid spread of coronavirus has left many hospitals facing unprecedented strains on resources, with limited capacity to care for critically ill patients.

To help researchers, healthcare leaders, and the public identify places with low capacity, some organizations are monitoring data on hospitals’ utilization and capacity rates.

Definitive Healthcare recently partnered with Esri to launch an interactive data platform that allows users to track US hospital bed capacity, as well as potential geographic areas of risk. The resource displays the location and number of licensed beds, staffed beds, ICU beds, and total bed utilization rates across the country.

Some institutions have also leveraged real-world data to design modeling tools that can help hospitals and health systems plan for critical care surges. A team from Penn Medicine developed a tool that would predict surges in clinical demand, as well as best- and worst-case scenarios of COVID-19-induced strain on hospital capacity.

Researchers used publicly available epidemiological data on COVID-19 and clinical outcomes data from multiple Penn hospitals to build the model.

“With close collaboration between the clinical and operational leaders of our health system and data science team, we were able to rapidly explore a range of scenarios based on published data from other regions of the world,” Penn researchers said.

RAND Corporation, a nonprofit research institution, recently created a similar model. Researchers developed an interactive tool that allows decisionmakers to estimate current care capacity and explore strategies for increasing it.

As part of the project, the group reviewed literature on past outbreaks and COVID-19 experiences, conducted surveys of frontline clinicians, and held virtual roundtables with emergency care providers.

“These critical care capacity estimates can inform cross-regional critical care resource sharing—from regions with less demand to those with more demand,” the RAND research team said.

“We encourage hospital leaders and regional and state officials to use this tool to examine critical care capacity creation strategies using assumptions based on data from their communities.”

Identifying high-risk patient populations

Healthcare leaders and investigators are also gathering and analyzing real-world data to determine who is most at risk during the COVID-19 pandemic.

At Medical Home Network (MHN), a Medicaid accountable care organization, care managers are identifying vulnerable individuals by finding out which patients are experiencing social isolation. MHN asks patients if they live alone, if they are homeless, and whether they have people who will help them if they get sick.

Additionally, the organization is leveraging AI and machine learning to identify which patients have a high risk of admission for COVID-19, or for unrelated complications from respiratory issues.

“After we cross those two lists of patients – those patients who are at risk for being admitted for respiratory failure or COVID-19, and those patients who are socially isolated – we know who to reach out to first,” Art Jones, MD, chief medical officer of MHN, told HealthITAnalytics.com.

Other institutions are collecting patient data to better understand COVID-19 risk factors. Researchers on the Healthy Nevada Project, a population health study combining genetic, clinical, and environmental data, is now incorporating COVID-19 data from consented participants.

In a 13-question online survey, study participants offered information about possible exposure or risks of COVID-19, including recent travel, attendance at large public events, and whether they are experiencing symptoms of the virus.

“The data that our participants have provided us, in less than a week, has allowed us to discover risk factors within communities and take action to live longer, healthier lives,” said Joseph Grzymski, PhD, an associate research professor at the Desert Research Institute (DRI), Chief Science Officer for Renown Health, and principal investigator of the Healthy Nevada Project.

State-level organizations are also utilizing real-world data to demonstrate the impact of COVID-19 on certain communities. The Illinois Department of Public Health (IDPH) is releasing COVID-19 cases by zip code, allowing people to see how the virus is affecting different areas of the state.

The information can help leaders recognize which geographic locations may need stricter interventions or more critical care resources.

Tracking intervention success, informing next steps

As the number of confirmed COVID-19 cases continues to climb, states across the country are adopting stringent intervention methods to curb the spread. Social distancing measures have become commonplace in US communities, and researchers have started to examine the potential impact of these approaches.

At the University of Texas Health Science Center at Houston (UTHealth), a group used an AI tool to discover that stricter, immediate interventions are needed to reduce the spread of coronavirus in the greater Houston area.

The researchers developed the model based on COVID-19 cases in China and Italy, and applied that model to 150 countries around the world. When the virus spread to the US, researchers first used the model at the state level and then the major metropolitan areas in Texas, including Houston.

“Although there are a lot of numbers and a lot of details, we saw two consistent patterns: earlier intervention was better, and more stringent intervention was better than less stringent,” said Eric Boerwinkle, PhD, dean and M. David Low Chair in Public Health at UTHealth School of Public Health.

Stanford University researchers took a broader approach. Investigators developed a data-driven tool that evaluates the possible outcomes of interventions like social distancing and quarantine. Rather than trying to map the exact dynamics of a particular location, the model shows possible trajectories under different hypothetical scenarios.

“We wanted to start a larger conversation about how our long-term response might look,” said Erin Mordecai, Stanford biologist. “We’re concerned about the potential for the disease to rapidly spread once we lift control measures.”

The modeling framework allows for different types, intensities, and durations of interventions to be implemented, and shows how these interventions impact the spread of the virus over time.

“In the future, we are considering additional interventions and scenarios including contact tracing with efficacy dependent on the testing capacity, fatality and hospitalization rates dependent on the age structure of a population, and fatality rates further dependent on hospital capacities,” Stanford researchers said.

As the COVID-19 situation continues to unfold, real-world data will increasingly help healthcare leaders make critical decisions to mitigate the impact of the virus. Tracking, controlling, and understanding coronavirus will largely depend on the industry’s ability to learn from past and current real-world information.