The goal of this prospective, multicenter, open-label, blinded end-point pragmatic study is to evaluate an artificial intelligence (AI)-augmented echocardiography screening approach for early detection of metabolic dysfunction associated steatotic liver disease (MASLD) and/or cirrhosis, in patients undergoing routine transthoracic echocardiograms (TTEs). The main question it aims to answer is to: 1. Evaluate notification responsiveness and rates of confirmatory testing for patients identified as high risk for having liver disease to determine whether optimized notifications increase timely confirmatory testing and treatment initiation versus standard of care assessment. 2. Compare time to diagnosis, treatment uptake, and clinical outcomes (hospitalizations, incident ASCVD, mortality) between cohorts identified as high risk by the AI algorithm and comparison groups to determine whether AI guided screening shortens time to diagnosis and increases appropriate treatment.
Study Type
INTERVENTIONAL
Allocation
NA
Purpose
DIAGNOSTIC
Masking
NONE
Enrollment
2,000
AI-generated notifications to clinicians about possible undiagnosed liver disease (MASLD and/or Cirrhosis) detected from Transthoracic Echocardiogram
Cedars-Sinai Medical Center
Los Angeles, California, United States
UCLA
Los Angeles, California, United States
Stanford Healthcare
Palo Alto, California, United States
Kaiser Permanente
Pleasanton, California, United States
Northwestern Medicine
Chicago, Illinois, United States
Massachusetts General Hospital
Boston, Massachusetts, United States
Positive Predictive Value (PPV) of the AI algorithm for detecting MASLD and/or cirrhosis confirmed within 12 months of AI identification.
Numerator: Participants with clinician-confirmed diagnosis of later stage MASLD and/or cirrhosis after confirmatory evaluation. Denominator: * Participants with positive AI screen who were enrolled and evaluated. * The intervention is the clinician referral or referral testing workflow. The clinicians ultimately have discretion to avoid further downstream testing if pretest probability is felt to be too low. If a clinician determines no further testing is warranted despite high risk assessment by AI, the participant will be classified as a false positive (still counted in the denominator).
Time frame: From enrollment to end of follow up at 1 year.
Time to diagnosis of MASLD/cirrhosis
Time (days) from AI identification to first confirmatory diagnosis
Time frame: Followed up to 24 months post notification.
Time to diagnosis for MASLD with F2 fibrosis or greater
Time (days) from AI identification to first confirmatory diagnosis
Time frame: Followed up to 24 months post notification.
Time to diagnosis for steatotic liver disease
Time (days) from AI identification to first confirmatory diagnosis
Time frame: Followed up to 24 months post notification.
Time to confirmatory imaging
Time (days) from AI identification
Time frame: Followed up to 24 months post notification.
Time to initiation of targeted treatment
Time (days) from AI identification
Time frame: Followed up to 24 months post notification.
All-cause mortality
Time (days) from AI identification
Time frame: Followed up to 24 months post notification.
All-cause hospitalization
Time (days) from AI identification
Time frame: Followed up to 24 months post notification.
Heart failure hospitalization
Time (days) from AI identification (Defined as admission with IV diuretics or elevated BNP)
Time frame: Followed up to 24 months post notification.
Cardiovascular hospitalization
Time (days) from AI identification for Cardiovascular hospitalization (defined by principal ICD9/10 code)
Time frame: Followed up to 24 months post notification.
Hepatic decompensation hospitalization
Time (days) from AI identification for hepatic decompensation hospitalization (defined by ascites, hepatic encephalopathy, variceal bleeding, hepatocellular carcinoma, or liver transplantation)
Time frame: Followed up to 24 months post notification.
New ASCVD diagnosis
Time (days) from AI identification
Time frame: Followed up to 24 months post notification.
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.