Clinical Decision Support Tool for Heart Failure

(LLM-GDMT Trial)

JW
Overseen ByJonathan W Cunningham, MD, MPH
Age: 18+
Sex: Any
Trial Phase: Academic
Sponsor: Brigham and Women's Hospital
Must be taking: Loop diuretics
No Placebo GroupAll trial participants will receive the active study treatment (no placebo)

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial tests a new tool designed to help doctors better manage treatments for people with heart failure. The LLM-GDMT Clinical Decision Support Tool uses data from medical records to suggest safe improvements to treatment. It sends helpful notes to heart doctors before appointments. Individuals diagnosed with heart failure in the last two years, who use certain heart medications or have specific heart test results, might be a good fit for this trial. As an unphased trial, this study offers a unique opportunity to contribute to innovative research that could enhance heart failure treatment for many patients.

Will I have to stop taking my current medications?

The trial does not specify if you need to stop taking your current medications. Treatment decisions, including medication changes, are made by your doctor and you.

What prior data suggests that this clinical decision support tool is safe for heart failure patients?

Research has shown that tools using large language models (LLMs) can assist doctors in making better treatment decisions for heart failure patients. These tools analyze patient information and suggest ways to improve care. Importantly, the tool in this trial provides advice but does not automatically change treatment. Doctors and patients make all decisions together.

Safety data from similar tools indicate they are generally safe to use. Reports of issues, such as hospital stays for heart failure, are rare, suggesting that using the tool is usually safe for patients. The tool helps doctors identify and avoid potential problems, enhancing treatment safety.12345

Why are researchers excited about this trial?

Researchers are excited about the LLM-GDMT Clinical Decision Support Tool because it introduces a cutting-edge way to assist doctors in managing heart failure. Unlike traditional treatments that rely solely on a clinician's expertise to decide on medications and therapies, this tool uses a large language model to analyze patient data from electronic health records. It generates personalized insights, helping doctors optimize treatment plans and consider safety precautions before patient visits. This advisory tool enhances decision-making without replacing the clinician's judgment, potentially leading to better patient outcomes.

What evidence suggests that this clinical decision support tool is effective for heart failure?

Research has shown that tools designed to assist doctors in decision-making can greatly enhance heart failure treatment. For example, one study found that adhering to recommended medical guidelines for heart failure significantly reduced hospital visits and increased survival rates. In this trial, one group of providers will use the LLM-GDMT Clinical Decision Support Tool, which employs advanced technology to analyze electronic health records and suggest improved treatment options. By advising doctors on medication changes, it helps them make more informed choices. Early evidence suggests that these AI tools can improve patient outcomes by supporting better care decisions. Meanwhile, the other group will continue with usual care without the tool during the initial evaluation phase.12346

Are You a Good Fit for This Trial?

This trial is for adult patients with heart failure who are seen in outpatient cardiology clinics at Mass General Brigham. It's designed to test if a new tool can help doctors follow heart failure treatment guidelines better.

Inclusion Criteria

At least one prior cardiology clinic visit in the MGB system within the past 2 years
I am currently using or have recently used water pills for heart failure.
Scheduled outpatient visit with a participating cardiology provider in an MGB outpatient cardiology clinic
See 4 more

Exclusion Criteria

History of heart transplant or presence of a left ventricular assist device
Systolic blood pressure <90 mmHg on the most recent recorded measurement
Heart rate <50 beats per minute on the most recent recorded measurement
See 4 more

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

Initial Evaluation

Providers are assigned to early implementation of the CDS tool or usual care. The CDS tool reviews EHR data and generates advisory messages for eligible encounters.

4 weeks
Messages delivered prior to scheduled visits

Follow-up

Participants are monitored for GDMT optimization and safety outcomes within 30 days of the index visit.

30 days

Extension

The CDS tool may be expanded to providers in the usual care arm as part of routine care after the initial evaluation phase.

What Are the Treatments Tested in This Trial?

Interventions

  • LLM-GDMT Clinical Decision Support Tool

Trial Overview

The study tests a clinical decision support tool that uses a large language model to analyze patient data and suggest improvements in heart failure treatments. Doctors receive these suggestions via their electronic health record system or email.

How Is the Trial Designed?

2

Treatment groups

Experimental Treatment

Active Control

Group I: Early ImplementationExperimental Treatment1 Intervention
Group II: Usual Care (Delayed Implementation)Active Control1 Intervention

Find a Clinic Near You

Who Is Running the Clinical Trial?

Brigham and Women's Hospital

Lead Sponsor

Trials
1,694
Recruited
14,790,000+

Citations

Large Language Model-Generated Messages to Improve ...

For eligible heart failure encounters, the CDS tool reviews existing electronic health record (EHR) data, including diagnoses, medications, ...

Clinical Decision Support Tools for Optimizing Guideline ...

Guideline-directed medical therapy (GDMT) for heart failure (HF) with a reduced ejection fraction confers a substantial benefit by reducing the ...

Artificial Intelligence for Cardiovascular Care in Action

The integration of AI in health systems has the potential to improve patient outcomes and enhance the efficiency of care delivery across diverse populations.

Applications of ChatGPT in Heart Failure Prevention, ...

AI algorithms have demonstrated the potential to improve HF care by supporting clinical decision-making, optimizing treatment allocation to ...

ADVOCATE Teaming Profiles

Core work focuses on scalable deployment of transparent human-centered AI for healthcare and defense, including clinical decision support tools developed for ...

Performance of Large Language Models in Interventional ...

Large language models (LLMs) have the potential to assist in complex decision making for interventional cardiology (IC). However, their comparative performance ...