Clinical Trial by Harvard Pilgrim Health Care Institute, HCA Healthcare, UCI Health and CDC Identifies Strategy to Rapidly Detect and Respond to Hospital Outbreaks Using Algorithm-Driven Technology

An automated tool that improves outbreak detection for hundreds of pathogens successfully served as an early warning system to find and respond to potential hospital outbreaks, as reported today in NEJM Evidence.

A Trial of Automated Outbreak Detection to Reduce Hospital Pathogen Spread

The large multi-state, real-world study was conducted in 82 hospitals and led by Harvard Pilgrim Health Care Institute, HCA Healthcare, University of California, Irvine (UCI) Health and the Centers for Disease Control and Prevention (CDC).

Contagious bacteria and other pathogens can spread in hospitals, increasing the risk of harmful infections in patients. While hospitals and health systems work to prevent infections and reduce the opportunity for outbreaks to occur, there is no standardized approach for detecting transmission. Early detection can lead to a rapid response that reduces the chance for outbreaks to occur.

"Despite significant progress in reducing healthcare-associated infection outbreaks, including of antimicrobial-resistant pathogens, they remain an industry challenge and can present as clusters that signal potential for transmission to patients,” said Joseph Perz, DrPH, MA, Senior Advisor for Public Health Programs in CDC’s Division of Healthcare Quality Promotion, and committee member for the CDC’s Council for Outbreak Response: Healthcare-Associated Infections. “The CLUSTER trial provides evidence that early detection powered by automation tools and quick action can prevent outbreaks from growing."

Most hospitals focus on a small number of antibiotic-resistant organisms and miss outbreaks from other pathogens. To address this, investigators developed an algorithm-driven statistical outbreak detection tool that used clinical laboratory data to provide real-time alerts to hospital infection prevention programs about potential transmission of over 100 bacterial and fungal species. This method relied on an automated review of organisms grown from patients’ clinical cultures and a statistical assessment of whether an increase was seen compared to prior experience. Detected increases triggered an automatic notification for hospital personnel to implement a response protocol to prevent additional cases.