Sharing Info Lowers Hospital Mortality Rates

Hospitals that had advanced IT systems and took part in clinical trials saw fewer deaths during the first year of the COVID-19 pandemic

Based on the research of Indranil Bardhan

There’s an unsung success story about the first year of the COVID-19 pandemic. At its outset, U.S. hospitals faced huge challenges, as influxes of patients strained resources, while doctors were uncertain about what treatments would work against the novel infection.

It’s remarkable that even before vaccines were widely deployed, hospitals were able to drastically reduce COVID-19 mortality rates, from 7.46% in April 2020 to 1.76% a year later.

New research from the McCombs School of Business helps explain why: better health information sharing among hospitals.

Indranil Bardhan, professor of information, risk, and operations management and Charles and Elizabeth Prothro Regents Chair in Health Care Management, found that counties with lower mortality rates tended to have two characteristics: They had hospitals that participated in clinical trials and used advanced health information technology.

Advanced health IT includes electronic health records (EHR) systems that can collect, organize, and share relevant health data, leading to more timely responses. Says Bardhan, “IT helps connect the dots across different hospital systems, or even across different networks of hospitals.”

With Christos Nicolaides of the University of Cyprus, Avinash Collis of Carnegie Mellon University, and Seth Benzell of Chapman University, Bardhan analyzed countywide data for hospitals’ IT capabilities, such as use of medical records. Data for individual hospitals were not available.

The researchers found that counties with advanced EHR use were more likely to be able to share data with other hospitals, to help them learn what treatments were working or not working against COVID-19. That sharing made a life-and-death difference.

• Counties with high IT ratings reduced mortality rates per capita 75% over the pandemic’s first year, compared with 47% for low-rated counties.
• If all hospitals in the study had higher IT levels, they would have seen 20,853 fewer deaths nationwide.

Learning From Trials and Errors

Another aspect of information sharing was participation in one or more of 4,968 clinical trials related to COVID-19, Bardhan says. Not only did clinical trials help determine effective treatments, but exchanging information about them with other institutions contributed to solutions.

“Even if it meant that certain trials were failures, it is the mindset and the ability of these hospitals to take risks. In terms of figuring out what kinds of interventions work, that is key,” he says.

He notes that COVID-19 mortality rates tended to come down even where trials were unsuccessful, suggesting that learning from trials and IT-assisted information sharing helped advance treatments.

Additionally, hospitals learned from hospitals in neighboring counties as well as their own, suggesting that learning from peers had a spillover effect.

By contrast, counties where hospitals did not participate in clinical trials and had low IT ratings fared the worst in COVID-19 mortality rates during the same period.

While researchers are still unpacking many of the implications of the pandemic, Bardhan says this paper’s lessons could prove useful during a future health crisis.

“Our research can help policymakers and decision-makers understand the value of information technology a little bit better,” he says. “Perhaps they can invest in the right kinds of information technology to help hospitals make faster and better decisions.”

“Learning From COVID-19: Clinical Trials, Health Information Technology, and Patient Mortality” is published in NPJ Health Systems.

Story by Omar Gallaga