How Yale New Haven Health uses eCART to support clinicians and patients

Yale New Haven (Conn.) Health has implemented an AI-powered clinical deterioration tool, eCART, across its seven hospitals — a move leaders say has both improved patient safety and reduced strain on clinicians.

The health system evaluated six different tools — three traditional and three AI-enhanced — on nearly 400,000 patients over three years before selecting eCART, Deborah Rhodes, MD, senior vice president of care signature and associate chief medical officer at Yale New Haven Health told Becker’s. She said the tool continuously monitors 97 clinical variables for every inpatient outside the ICU, identifying not just abnormal values but patterns and interactions that can signal a patient’s decline hours before it becomes critical.

“Most tools are set to fire around fixed parameters, so they over-alert on stable patients who just run low or high,” said Dr. Rhodes. “By being able to identify patterns in individual patients that signal a change or interaction, eCART is able to have this higher level of precision relative to other tools.”

Dr. Rhodes noted that one of the tool’s biggest advantages is the lead time it provides. On average, eCART identifies signs of patient deterioration about four hours before a critical event, giving clinicians time to intervene. She said that advance notice has become central to how the system is learning to improve its response to patient decline.

“We are saving lives, and we have the data to show that,” Dr. Rhodes said.

Since going live systemwide more than a year ago, Yale New Haven Health has reported a 13% relative decrease in mortality — a 1.2% absolute reduction — with especially notable gains in sepsis-related deaths. 

Unlike some predictive tools that overwhelm clinicians with alerts, eCART typically flags no more than 20 to 30 patients at a time in even the largest hospital. When a patient’s score rises, nurses follow a structured pathway beginning with rechecking vital signs and, if needed, calling in a rapid response team. Each unit receives weekly feedback on how quickly staff respond and complete the pathway, which Dr. Rhodes said has helped standardize care across the system.

“Most new things you ask teams to do are met with resistance because they’re already overwhelmed. What we’ve found with eCART is the opposite — rather than adding to their burden, it helps them focus on the patients who most need attention. The nurses have just embraced this,” she said.

Dr. Rhodes credits not only the technology but the system built around it. 

“AI is the heightened signal. Augmented intelligence is when we build the right processes, structures and teamwork around the tool to really take it to its full potential,” she said. “The tool itself is only step one. It’s the steps we built around it that got us to this state.”

Looking ahead, Yale New Haven Health is preparing to expand eCART into its children’s hospital for patients ages 2 to 18. While FDA approval covers only adult populations, the health system’s internal data showed statistically significant performance in pediatric patients as well.

The post How Yale New Haven Health uses eCART to support clinicians and patients appeared first on Becker’s Hospital Review | Healthcare News & Analysis.