Sepsis is a serious condition that occurs when the body’s immune system has an extreme response to an infection. It can lead to organ failure, tissue damage and death. According to a recent study, an artificial intelligence (AI) tool can help detect sepsis risk in patients and potentially save lives.
How COMPOSER works
The study, published in the journal npj Digital Medicine, was conducted by researchers at the University of California (UC) San Diego School of Medicine. They developed an AI algorithm called COMPOSER, which stands for COntinuous Multimodal Prediction of Sepsis using REcurrent neural networks.
COMPOSER works by continuously monitoring more than 150 different patient variables that could be linked to sepsis, such as lab results, vital signs, current medications, demographics and medical history. It uses a deep-learning model, which is a type of AI that can learn from complex and multiple data sources.
If COMPOSER detects a high risk of sepsis infection in a patient, it alerts the nursing staff via the hospital’s electronic health record. The nursing team then reviews the alert with the physician and decides on the appropriate treatment plan.
The benefits of COMPOSER
The researchers tested COMPOSER in the emergency departments at UC San Diego Medical Center and at Jacobs Medical Center. They compared more than 6,000 patient admissions before and after COMPOSER was deployed.
They found that COMPOSER resulted in a 17% reduction in mortality among patients who developed sepsis. They also found that COMPOSER reduced the time to sepsis diagnosis by 4 hours and the time to antibiotic administration by 3 hours.
The researchers said that COMPOSER is the first AI tool to show improvement in patient outcomes by using a deep-learning model to predict sepsis risk. They said that COMPOSER can help prevent sepsis complications and save lives by providing early and accurate detection of sepsis.
The future of COMPOSER
The researchers plan to expand the use of COMPOSER to other hospitals and health care settings. They also hope to improve COMPOSER by adding more data sources, such as genomic and microbiome data, to enhance its predictive power.
They said that COMPOSER is an example of how AI can improve health care delivery and quality by providing decision support tools for clinicians. They said that COMPOSER can complement human judgment and expertise by identifying complex and subtle patterns of sepsis risk.