We use cookies to understand how you use our site and to improve your experience. This includes personalizing content and advertising. To learn more, click here. By continuing to use our site, you accept our use of cookies. Cookie Policy.

HospiMedica

Download Mobile App
Recent News AI Critical Care Surgical Techniques Patient Care Health IT Point of Care Business Focus

AI-Based System to Recommend Clinical Treatments for Sepsis Patients in ICU

By HospiMedica International staff writers
Posted on 01 May 2023

In an intensive care unit (ICU), clinicians must make complex decisions quickly and accurately, constantly monitoring critically ill or unstable patients. Researchers have now developed an artificial intelligence (AI)-based system to aid physicians in making decisions within the ICU.

Researchers from Carnegie Mellon University (Pittsburgh, PA, USA) collaborated with physicians and other researchers to explore whether AI could assist in decision-making and whether clinicians would trust such support. The team provided 24 ICU physicians with access to an AI-based tool designed to aid decision-making and found that most integrated the assistance into some of their decisions. Using the 2018 AI Clinician model, they developed an interactive clinical decision support (CDS) interface—named AI Clinician Explorer—that offers recommendations for treating sepsis. The model was trained on a data set of over 18,000 patients who met standard diagnostic criteria for sepsis during their ICU stays. The system allows clinical experts to filter and search for patients in the data set, visualize their disease trajectories, and compare the model predictions to actual bedside treatment decisions.


Researchers have developed an AI-based system to recommend clinical treatments (Photo courtesy of Freepik)
Researchers have developed an AI-based system to recommend clinical treatments (Photo courtesy of Freepik)

The team conducted a think-aloud study with 24 ICU clinicians experienced in sepsis treatment, having them use a simplified AI Clinician Explorer interface to assess and make treatment decisions for four simulated patient cases. The team observed four common behaviors among the clinicians: ignore, rely, consider, and negotiate. The "ignore" group disregarded the AI's influence, while the "rely" group consistently accepted at least part of the AI's input. The "consider" group contemplated the AI's recommendation before accepting or rejecting it. Most participants belonged to the "negotiate" group, accepting individual aspects of the recommendations in at least one decision, but not all.

The team found the results surprising and gained insights on how to improve the AI Clinician Explorer. Clinicians expressed concerns about the AI lacking access to more holistic data, such as the patient's general appearance, and were skeptical when the AI made recommendations contrary to their training. The research aims not to replace or replicate clinician decision-making, but to use AI to reveal patterns that may have been previously overlooked in patient outcomes.

"It feels like clinicians are excited about the potential for AI to help them, but they might not be familiar with how these AI tools would work. So it's really interesting to bring these systems to them," said Venkatesh Sivaraman, a Ph.D. student in the HCII and member of the research team. "There are a lot of diseases, like sepsis, that might present very differently for each patient, and the best course of action might be different depending on that. It's impossible for any one human to amass all that knowledge to know how to do things best in every situation. So maybe AI can nudge them in a direction they hadn't considered or help validate what they consider the best course of action."


Gold Member
12-Channel ECG
CM1200B
Gold Member
SARS‑CoV‑2/Flu A/Flu B/RSV Sample-To-Answer Test
SARS‑CoV‑2/Flu A/Flu B/RSV Cartridge (CE-IVD)
Silver Member
Wireless Mobile ECG Recorder
NR-1207-3/NR-1207-E
New
Eye-Level Digital Physician Scale
icon

Latest AI News

AI-Powered Algorithm to Revolutionize Detection of Atrial Fibrillation