Study: AI-powered alerts improve patient outcomes

A study done at Mount Sinai Hospital and reported in Critical Care Medicine found that AI-generated alerts significantly improve patient care and outcomes.

The study found that hospitalized patients whose care teams received AI-generated alerts about adverse changes in their health were 43% more likely to receive timely escalated care and significantly less likely to die. The research was led by Dr. Matthew A. Levin, professor of anesthesiology, perioperative and pain medicine, and genetics and genomic sciences at Icahn Mount Sinai.

“We wanted to see if quick alerts made by AI and machine learning, trained on many different types of patient data, could help reduce both how often patients need intensive care and their chances of dying in the hospital,” said Levin in a statement. “Traditionally, we have relied on older manual methods such as the Modified Early Warning Score to predict clinical deterioration. However, our study shows automated machine learning algorithm scores that trigger evaluation by the provider can outperform these earlier methods in accurately predicting this decline. Importantly, it allows for earlier intervention, which could save more lives.”

Study details

The non-randomized, prospective study included 2,740 adult patients admitted to four medical-surgical units at The Mount Sinai Hospital. Patients were divided into two groups: one receiving real-time alerts sent to their nurses, physicians, or a rapid response team, and another where alerts were generated but not sent. In units with suppressed alerts, urgent interventions were based on standard deterioration criteria.

Key findings in the intervention group indicated that patients were more likely to receive early medications to support heart and circulation and were less likely to die within 30 days.

“Our research shows that real-time alerts using machine learning can substantially improve patient outcomes,” said David L. Reich, president of The Mount Sinai Hospital and Mount Sinai Queens, in a statement. “These models are accurate and timely aids to clinical decision-making that help us bring the right team to the right patient at the right time. We think of these as ‘augmented intelligence’ tools that speed in-person clinical evaluations by our physicians and nurses and prompt the treatments that keep our patients safer. These are key steps toward the goal of becoming a learning health system.”

Although the study was cut short due to the COVID-19 pandemic, the algorithm has since been implemented in all stepdown units at The Mount Sinai Hospital. Stepdown units are specialized areas for patients who require close monitoring but are not in critical condition. Intensive care physicians now review the 15 patients with the highest prediction scores daily and recommend treatments. The algorithm, which is continually retrained with more patient data, improves in accuracy through reinforcement learning.

This clinical deterioration algorithm is one of 15 AI-based clinical decision support tools developed and deployed by the researchers throughout the Mount Sinai Health System, illustrating the integration of advanced AI solutions in health care to enhance patient safety and outcomes.