Multicenter Development and Validation of a Multimodal Deep Learning Model to Predict Moderate to Severe AKI.
Jay L Koyner, Jennie Martin, Kyle A Carey...
https://pubmed.ncbi.nlm.nih.gov/40232856Actively Recruiting
Led by University of Chicago · Updated on 2025-09-12
800
Participants Needed
2
Research Sites
52 weeks
Total Duration
U
University of Chicago
Lead Sponsor
N
National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)
Collaborating Sponsor
Researchers are evaluating how combining kidney-related biomarkers from blood and urine with a new computer-based risk score can better identify hospitalized patients at high risk for severe acute kidney injury (AKI). This observational study focuses on patients who have a high risk of developing Stage 2 AKI, using a natural language processing (NLP) based AKI risk algorithm that analyzes electronic health record data from labs, vital signs, clinical notes, and test reports. The study is conducted at the University of Chicago Medical Center and the University of Wisconsin Hospital. Patients identified as high risk (top 10% risk score) will be enrolled and have blood and urine samples collected for biomarker testing over three days. The study plans to recruit two groups of 400 patients each across the two hospitals. The first group will help researchers determine if adding biomarkers to the electronic risk score improves prediction of Stage 2 AKI and other outcomes. The findings will then be validated in the second group. The risk score software is used as a noninvasive device but will not directly impact clinical care during the study. Participants will be admitted patients at inpatient wards, intermediate care, or ICU, and will be monitored for the development of KDIGO Stage 2 AKI within seven days of enrollment. Researchers will collect clinical data continuously and measure outcomes including Stage 3 AKI, need for kidney dialysis, mortality during hospitalization, and major adverse kidney events over 90 days. The study involves consent procedures and excludes patients with certain kidney conditions, prior AKI episodes during the admission, or those unable to consent in English. The total study participation period includes the initial 3-day biomarker collection and follow-up for up to 90 days.
CONDITIONS
Combining Biomarkers and Electronic Risk Scores to Predict AKI in Hospitalized Patients
You may qualify if you...
You will not qualify if you...
Complete this quick 3-step screening to check your eligibility
Duration - 2 to 4 weeks
Participants are screened for eligibility to participate in the trial.
Duration - 3 days
Participants identified as high risk based on their AKI risk score will have blood and urine samples collected to evaluate biomarkers.
Daily visits for 3 days
Duration - Up to 90 days
Participants will be monitored for development of Acute Kidney Injury (AKI) and other clinical outcomes up to 90 days after enrollment.
Follow-up assessments within 7 days and at 90 days post-enrollment
Total: 2 locations
1
University of Chicago Medical Center
Chicago, Illinois, United States, 60637
Actively Recruiting
2
University of Wisconsin Hospital
Madison, Wisconsin, United States, 53792
Actively Recruiting
J
Jay Koyner, MD
Study Type
OBSERVATIONAL
Masking
N/A
Allocation
N/A
Model
N/A
Primary Purpose
N/A
Number of Arms
1
Have more questions? Get in touch with our team for quick support
Explore thousands of other clinical trials that might be a better match.
Sign up to get personalized trial recommendations delivered to your inbox.
Already have an account? Log in here
Jay L Koyner, Jennie Martin, Kyle A Carey...
https://pubmed.ncbi.nlm.nih.gov/40232856