The tool aims to predict a patient's chances of dying within 5-90 days of admission, helping care teams to decide when and how to integrate ACP into care management.
OSF Healthcare is using AI to help doctors and nurses integrate end-of-life discussions into care management plans.
A research team at the Illinois health system led by OSF Senior Fellow for Innovation Jonathan Handler, MD, tested an AI model that predicts the likelihood of a patient's death five to 90 days after admission. That information is then used by care teams to decide when to begin advanced care planning (ACP) for patients and their families.
The tool could help health systems improve care for a large number of patients. Surveys estimate only 22% of Americans have documented their end-of-life wishes. ACP can reduce the use of complex or intensive treatments at the end of life, thus reducing the cost and length of hospital stays and the amount of anguish placed on family members.
"Although experts agree on the importance of ACPs, clinicians cite time constraints and poor communication with other providers as barriers to having end-of-life discussions," Handler and his team wrote in a recently published study. Reduced access to healthcare in mixed-rurality populations may make ACP even more unlikely. Due to these barriers, many patients do not have documented preferences at the end-of-life and therefore do not achieve what has been termed an 'ideal death.'"
The researchers tested the tool on a dataset of more than 75,000 inpatient visits both before and during the pandemic, ensuring that the tool holds up over time and is equitable across genders, races and ethnicities, and against rural and socioeconomic factors. According to the study, the model helped to identify more than half of patients within the 5- to 90-day range.
Handler and his team also compared their model against one developed two years ago at NYU Langone Health, which was used to predict mortality rates within 60 days. Other health systems that have tested this type of tool include Penn Medicine, Stanford Medicine, and Duke Health.
"We sought a model to predict post-inpatient mortality to meet a different need – to help prioritize and encourage timely ACP conversations during an inpatient stay," the OSF team wrote. "The model’s intended use is to predict mortality soon after the length of an average inpatient stay. Therefore, the 5-to-90-day window was chosen to: 1) begin after the average 4-day length of an inpatient stay, 2) allow at least 4 days for an ACP if the inpatient stay is longer than average, and 3) create enough urgency to stimulate the ACP."
In their conclusion, Handler and his colleagues say their model holds up well over time and can help to "consistently and equitably help prioritize patients likely to benefit in the near-term from theses crucial conversations."
Eric Wicklund is the associate content manager and senior editor for Innovation, Technology, Telehealth, Supply Chain and Pharma for HealthLeaders.
Only about 22% of Americans have documented end-of-life wishes for healthcare.
Advanced care plans (ACPs) are important in helping care teams determine the level of clinical care, which in turn affects the cost and length of the hospital stay and the ability to counsel patients and their families.
Health systems like OSF are using AI tools to better predict a patient's last days, enabling clinicians to integrate ACP into the care plan at the right time.